Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at SciVerse ScienceDirect

Optics and Lasers in Engineering

journal homepage: www.elsevier.com/locate/optlaseng

Review

The biospeckle method for the investigation of agricultural

crops: A review
Artur Zdunek n, Anna Adamiak, Piotr M. Pieczywek, Andrzej Kurenda
Institute of Agrophysics, Polish Academy of Sciences, Doswiadczalna 4, 20-290 Lublin, Poland

art ic l e i nf o a b s t r a c t

Article history: Biospeckle is a nondestructive method for the evaluation of living objects. It has been applied to
Received 3 April 2013 medicine, agriculture and microbiology for monitoring processes related to the movement of material
Received in revised form particles. Recently, this method is extensively used for evaluation of quality of agricultural crops. In the
13 June 2013
case of botanical materials, the sources of apparent biospeckle activity are the Brownian motions and
Accepted 28 June 2013
biological processes such as cyclosis, growth, transport, etc. Several different applications have been
shown to monitor aging and maturation of samples, organ development and the detection and
Keywords: development of defects and diseases. This review will focus on three aspects: on the image analysis
Biospeckle and mathematical methods for biospeckle activity evaluation, on published applications to botanical
Fruit
samples, with special attention to agricultural crops, and on interpretation of the phenomena from a
Quality
biological point of view.
Nondestructive
& 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction: biospeckle phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2. Evaluation of biospeckle activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1. Global measures of speckle activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1.1. Spatial and temporal contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1.2. Time history of the speckle pattern (THSP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.3. Spatial-temporal speckle correlation technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2. Spatial analysis of biospeckle activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1. Fujii's method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2. Generalized difference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.3. Laser speckle contrast analysis (LASCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.4. Motion history image (MHI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.5. Analyses in spectral domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3. Applications of the biospeckle-based methods in practical evaluation of agricultural crops. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1. Monitoring of aging and shelf life of botanical samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2. Relation of biospeckle activity with biochemical processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3. Defects and diseases inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4. Plant development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4. Biological bases of the biospeckle phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1. Introduction: biospeckle phenomena

The biospeckle phenomenon occurs when a biological material

is illuminated by a coherent light (laser light). The light is back-
n
Corresponding author. Tel.: +48 81 7445061; fax: +48 81 7445067. scattered from an optically rough surface and due to interferation,
E-mail address: [email protected] (A. Zdunek). bright and dark areas appear on an observation plane (Fig. 1).

0143-8166/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.optlaseng.2013.06.017

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
2 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

When the investigated object is static in time, the pattern is also light, which might be a diode laser, a detector such as a CCD
frozen. In the case of a biological material, first, light can usually camera with a lens and a PC with a frame grabber that is able to
penetrate tissue and can be backscattered either from the surface record a set of images with a constant time lag. From the time
or by internal inhomogeneity. Second, the material, when living, is series images, the biospeckle activity is estimated using various
not physically and chemically stable in time and space. This causes mathematical methods. These features make this method attrac-
the speckle pattern to also comprise a dynamic component. Draijer tive to many applications, which require fast and nondestructive
et al. suggested that this dynamic behavior is mainly caused by sampling.
Doppler shifts of the light that interacts with the moving particles [1]. This review consists of three main paragraphs presenting the
Particularly for biological media, the phenomenon of dynamic speckle current stage in development of biospeckle technique and its
pattern has been called as the biospeckle. application, particularly for plant tissue evaluation. Since the
The apparent activity of biospeckles is the result of physical device is very simple and therefore most often is not a key issue,
movement of particles inside cells and is affected by the variation first we will focus on the mathematical methods for measuring
of the absorption of light by tissue pigments. Therefore, the biospeckle activity and will show different approaches, including
activity of biospeckles can provide information about various their pros and cons. Then the current stage in various applications
living processes occurring inside a cell. Moreover, the method is for botanical samples will be presented. The final section will
considered to be a nondestructive one because no visible results of provide insight into interpretation of the phenomena for plant
tissue interaction with light have been reported up to now when tissue from a biological point of view.
low power lasers have been used. A typical setup for biospeckle
measurement is very simple (Fig. 2). It requires an expanded laser
2. Evaluation of biospeckle activity

2.1. Global measures of speckle activity

2.1.1. Spatial and temporal contrast

Assuming that the pattern produced by the light scattered by
the biological medium is composed of two components, one
produced by stationary scatterers and one by moving scatterers,
Briers concluded that, for long enough exposures or integration
times, the time-averaged speckle pattern of stationary scatterers is
constant at all points Briers [2]. Therefore, he expressed the
fluctuations of intensity of the speckle pattern as the ratio of the
mean intensity of the light from the moving scatterers to the total
intensity of the scattered light. This value, denoted by ρ and given
in terms of the variance of the speckle patterns, is expressed by
the following equation:
s
ρ ¼ 1 ð1Þ
〈I〉
where s2 is the spatial variance of the time-averaged speckle
Fig. 1. A snapshot (a speckle pattern) from a movie of biospeckles recorded on an
intact apple. The speckle pattern is the result of interferation of backscattered pattern and 〈I〉 is the mean intensity of the pattern. This measure is
coherent light from the surface and internal structure. The biospeckles contain a called the spatial contrast and can be calculated from a single-
static component from non-moving tissue elements and a dynamic component exposure image speckle pattern. On images obtained using this
from non-static tissue elements. To watch the biospeckle movie we recommend technique the areas with high speckle fluctuations are blurred
reference to Zdunek and Cybulska [17].
depending on the fluctuation velocity and therefore the spatial
contrast of these areas is reduced. In areas where there are no
fluctuations, the spatial contrast remains high [3]. Thus, the
integration time is the key factor in single-exposure photography
of biospeckle patterns and should be chosen depending on the rate
of observed fluctuations. Development of imaging techniques
using a CCD camera led to an alternative method of time averaging
where the ‘average’ frame is calculated from a set of collected
frames. Moreover, it was shown that in the case of a set of sampled
frames it is possible to perform direct application of temporal
statistics, where the process of time averaging is replaced by
spatial averaging [4]. In this approach the first-order temporal
statistics of fluctuations in speckle intensity are calculated at each
point separately. This alternative measure, called temporal con-
Fig. 2. Schematic view of the typical setup used for biospeckle measurement.
trast, is expressed as:
The system must contain a coherent laser light with optics to illuminate the object " #1=2
surface evenly. A CCD camera is the typical detector used. The objective of the CCD 〈s2t ðx; yÞx;y 〉
ρ ¼ 1 1 ð2Þ
camera is usually strongly defocused. The aperture of the objective determines 〈I〉2
speckle size d according to formula d ¼ 1.22(λz/D), where λ is laser wavelength, z is
the distance of observation and D is the diameter of the illuminated area. Larger where st2 (x,y) is the temporal variance of the intensity fluctua-
static speckles from the surface depend on the light incidence angle, whereas tions at point (x,y) of the speckle pattern and 〈…〉x,y indicates
smaller dynamic speckles resulting from interior of tissue are angle independent.
The final protocol for biospeckle measurement should consider laser wavelength,
spatial averaging. 〈I〉 is the mean intensity of the pattern in the
the distance and angle between samples and the detector, polarization and light temporal and spatial domain. Although spatial and temporal
intensity and the detector objective aperture. contrasts are velocity-dependent values, they do not reflect

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

information on the velocity of moving speckles or the frequency The IM is defined as the sum of the matrix values times the
components of intensity fluctuations. Instead, they estimate the squared distance from the principal diagonal. It is expected that
relative magnitude of speckle fluctuations [5]. Both methods use a larger IM relates to higher biospeckle activity. Experiments with
first-order statistics of the speckle fluctuations, giving information a simulated THSP showed that the calculated IM decreases with
only about the relative contribution of moving scatterers to a increasing length of the speckles and with speckle length above
whole fluctuating speckle pattern [6]. one third of the window size the IM reaches its saturation state.
The inertia moment showed sensitivity mainly to the mean value
of speckle length. However, tests have indicated that the presence
2.1.2. Time history of the speckle pattern (THSP) of noise, which produces sudden jumps in intensity, might cause
The THSP [7] is a matrix composed of a number of successive interference and result in an increase in IM values [8,11].
images of dynamic biospeckle phenomena. The THSP is created The absolute value of differences (AVD) was proposed as an
using the same column extracted from each image and placed side alternative to the routine IM method [10]. Suspecting that the square
by side. The set of columns is arranged sequentially in chronolo- operation carried out in the IM method can emphasize biospeckle
gical order. The width of the THSP is equal to the number of activity changes in a distorted way, this new approach was based on
images used and represents the time scale of biospeckle activity. the calculation of the absolute value of the distance from the principal
The THSP is composed of pixels in gray scale, arranged in diagonal. Analysis carried out by the AVD showed better results in
horizontal lines that show the time evolution of biospeckle some cases of biospeckle activity, especially when the THSP matrix
intensity fluctuations (Fig. 3a). contained no data at high frequencies.
Arizaga et al. suggested an approach to characterize speckle time Passoni et al. proposed wavelet entropy (WE) as the descriptor
evolution based on the inertia moment (IM) of the co-occurrence of the dynamic speckle phenomenon [12]. Wavelet entropy was
matrix of the THSP [8]. The co-occurrence matrix [9,10] is built used as a measure of order and disorder in a dynamical multi-
from the THSP image and shows sample activity as a spread of frequency signal of speckle activity. Each row of the THSP was
non-zero values outside the principal diagonal. For low activity, considered as an individual time series. A single image descriptor
the values of the matrix are concentrated around the diagonal, was calculated as the mean value of WE from all rows of the THSP
while for high activity the matrix resembles a cloud. An image of image. WE showed better agreement with experimental data than
the co-occurrence matrix in false color-scale is shown in Fig. 3b. the inertia moment for changes in biospeckle pattern relating

Fig. 3. Two examples presenting low (a and b) and high (c and d) biospeckle activity analysed using time history speckle pattern (THSP) method. The THSP matrixes
presented in (a)–low biospeckle activity and (c)–high biospeckle activity has been constructed from 512 vertical lines recorded to show each pixel intensity history on the
horizontal axis. The pixels in the co-occurrence matrix (COij), presented in (b)–low biospeckle activity and (d)–high biospeckle activity, represent the number of occurrences
of pixel intensity j after i. Since usually a CCD camera uses an 8-bit system, the CO matrix has dimensions of 256  256. A normalized inertia moment (IM) is calculated,
which is proportional to the spread of pixels in the CO matrix along its diagonal. A larger IM reflects higher biospeckle activity.

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
4 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

to low frequencies [11,13–15]. This technique permitted both 2.2. Spatial analysis of biospeckle activity
qualitative and quantitative estimations, provided in shorter time
and requiring less data compared to other methods of THSP 2.2.1. Fujii's method
analysis. At first, Fujii's method [22,23] was applied in blood flow
observations using time varying laser speckles. This method is
2.1.3. Spatial-temporal speckle correlation technique based on calculations of weighted sums of the differences between
This technique is based on the correlation analysis of two or two consecutive elements of time sequence of intensities of each
more speckle patterns, where one is considered as an image of a pixel. The value of Fujii's index for a single pixel is defined as:
reference state. The reference pattern and the patterns of subse- N jI k ðx; yÞI kþ1 ðx; yÞj
quent object states are divided into an equal number of regions Fðx; yÞ ¼ ∑ ð3Þ
k ¼ 1 I k ðx; yÞ þ I kþ1 ðx; yÞ
and then cross-correlation of each pair of respective fragments is
calculated. The calculation of the cross-correlation coefficients for where k is the image index from the sequence k ¼1…N and Ik is
a series of speckle pattern subimages recorded in the given the intensity value of a pixel with x and y coordinates. The image
temporal order allows the temporal dependencies of these coeffi- composed of integrated F(x,y) values shows regions of high and
cients to be received as functions of the biospeckle pattern move- low speckle intensity with preservation of the contours of the
ment speed [16]. Each such dependency is equivalent to temporal tested object (Fig. 5a). The process of weighting the differences by
degradation of a correlation peak. For a homogenous biospeckle the inverse of the sum of two subsequent intensity values results
pattern the fluctuations of speckle intensity are equal in each in a nonlinear response that emphasizes both the large differences
region, therefore the correlation peaks in all locations show similar as well as the small ones that involve values from the limits of the
degrees of biospeckle pattern degradation in time. In this case a dynamic range. However, Fujii's method has important drawbacks.
grid of peaks can be replaced by one averaged value equal to the As can be seen, the same difference in intensity value can be dealt
value of correlation calculated between two images. with in completely different ways. For instance, the transition
Biospeckle activity evolution can be evaluated by calculating from gray level 0 to 1 corresponds to the maximum value of Fujii's
temporal changes in the correlation coefficient Ckτ, where k is the index, while for a transition from 254 to 255 it takes the lowest
frame number and τ is the lag time between frames (Fig. 4). possible non-zero value. This might lead to the presence of false
In practice, Ckτ is calculated as the correlation coefficient of the activity in darker regions of the biospeckle activity map generated
data matrix, consisting of intensities of pixels, of the first frame by noise. A possible approach for overcoming these drawbacks is
with the data matrixes of the following frames of biospeckles. to use the frequency decomposition carried out by the wavelet
Typically, the lag time is limited by the detector capabilities and transform [15,24,25].
most often is about 1/15 s. A more pronounced decrease in Ckτ
(larger decorrelation) reflects higher biospeckle activity, therefore
the value 1  Ckτ is used for a more meaningful representation. 2.2.2. Generalized difference
To simplify the analysis in practical application it is enough to take The Generalized Difference (GD) method was introduced as an
just two snapshots of biospeckles with a lag of several seconds and alternative to Fuji's approach [26]. In the basic version of this
calculate the correlation coefficient between them [16–21]. This algorithm the weighting process is eliminated. The GD is defined
“longer” lag time should be found for a certain system and a as the cumulative sum of absolute values of the differences between
material should be investigated as a compromise between the the pixel intensities among all frames and can be written as:
time of measurement and the biospeckle decorrelation ratio.  
N1 N
 
GDðx; yÞ ¼ ∑ ∑ I k ðx; yÞI l ðx; yÞ ð4Þ
k ¼ 1 l ¼ kþ1

where k and l are the frame indexes and Ik is the intensity value of
a pixel with x and y coordinates. Unlike Fujii's method, the
algorithm of GD includes the differences between nonconsecutive
frames. The minimum possible value of GD is obtained when there
are no differences between pixel intensities among all the exam-
ined time scales (or in other words, the differences are equal to
zero). When half of the pixel intensity values from sequence
k¼ 1…N is equal to the upper bound of the dynamic range and
the second half possess the lower bound value, the calculated GD
reaches its maximum. However, due to the calculation of differ-
ences between all possible pairs of frames, the order of appearance
is not taken into account Therefore, this procedure generates a loss
of the temporal information of pixel intensity activity and shows
only the spread of values without any information about the
frequency of transitions. The contours of the examined object on
the resulting map of activity are less visible when compared to the
Fujii's method. The GD algorithm, when implemented directly, is
Fig. 4. Decrease of correlation coefficient C in time, calculated between the first also more computationally time consuming. There is also other
and consecutive frames of a biospeckle movie for a tomato. Biospeckle activity BA
variant of this method, where absolute values are replaced by
can be calculated as 1-C at a chosen time of biospeckle evolution. For most fruits
there are 4 s of evolution, however this depends on laser quality (coherence). squared values. This measure, denoted by GDn, is defined as
For example, for a low cost diode laser, as was used by Zdunek and Herppich [18], follows:
the evolution time was increased up to 14 s. This method is therefore very
convenient for future industrial application because it requires taking just two N1 N
snapshots with a lag of 4 s and allows determination of C between the two GDn ðx; yÞ ¼ ∑ ∑ ðI k ðx; yÞI l ðx; yÞÞ2 ð5Þ
matrixes. k ¼ 1 l ¼ kþ1

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

Fig. 5. Real view (a), biospeckle image and (b)spatial biospeckle activity of the apple after local mechanical treatment in the center of apple, visualized by Fujii's (c) and
LASCA (d) method. This example shows that biospeckle imaging allows the distinction of intact and damaged areas (dark spot in the center of apple) even when fruit damage
is still externally invisible. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).

These two approaches are not directly comparable, but have method are “laser speckle contrast imaging”, or “laser speckle
qualitative similarities [27]. Images of biospeckle activity gener- imaging” [30–32]. In general, the LASCA algorithm is based on
ated by the GDn method are characterized by a higher contrast calculations of spatial or temporal contrast over a local, square
compared to GD. It has been shown that GDn has a straight window of MxM pixels. The window is moved along the speckle
relationship with variance, therefore due to similarities between pattern and the calculated contrast value is assigned to the central
GD and GDn there is also a qualitative relationship between GD and pixel. The smaller the window, the lower the statistical validity,
variance. The loss of temporal data resulting from this fact was while a larger window reduces the effective resolution of the
eliminated in the WGD algorithm [26] by introducing to Eq. 4 an image. This method has many variations. In the most basic form
additional weight, which value varied along each summation of LASCA, the contrast in each location is calculated from a
depending on time scale. The WGD is defined as: single image of the speckle pattern using a moving window. For
N1 N   spatially derived contrast, LASCA is computed from set of collected
 
GDðx; yÞ ¼ ∑ ∑ I k ðx; yÞI l ðx; yÞwp ð6Þ frames [33]. The calculations are performed for each frame
k ¼ 1 l ¼ kþ1 separately, as in the case of single-exposure photography. After
where p ¼l  k and stands for the temporal distance from reference the contrast values are computed, they are averaged at each
frame k. The weight values vary with temporal distance p, location according to the present number of frames. The tempo-
emphasizing fast or slow variations in speckle intensity. rally derived contrast is calculated for each pixel separately using
The Fujii's and GD methods have one drawback in common that data from all frames [30,34] and then smoothing by the moving
the resulting matrix describes the speckle pattern activity as a window is employed.
whole during the observations interval, but misses the evolution Biospeckle pattern mapping can be carried out using over-
of the activity. It is not the case in a similar method to them, a lapping or contiguous windows. Using overlapping windows
temporal difference method which bases on subtraction of two results in smoother images, however as in case of contiguous
separated by a time interval speckle images [28]. This relatively windows, the effective resolution of the final image is reduced by a
simple and modest computational costs method allows detection factor of M. Window size is adjusted according to the size of the
whether the speckle structure has changed or not with addition of speckles. The most commonly used window sizes are 3  3, 5  5
obvious physical meaning of output pattern. or 7  7. Larger windows are not used due to too a high reduction
in image resolution. Due to the relatively fast computation
of LASCA, this method is applied for real-time monitoring. The
2.2.3. Laser speckle contrast analysis (LASCA) biggest drawback of LASCA in relation to other methods is the loss
LASCA is a technique that has been developed using a modified of resolution [24,35]. Similarly to the Fujii's method, the
Briers contrast method [2–4,29] (Fig. 5d). Other terms used for this same range of fluctuations in intensity (expressed by variance or

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
6 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

standard deviation) will give a different contrast value in darker 3.1. Monitoring of aging and shelf life of botanical samples
regions than in more illuminated ones.
In the case of plant tissue, biospeckle fluctuations are usually
not a result of well-defined fluxes of matter, such as those
2.2.4. Motion history image (MHI) observed during blood flow. Therefore these case studies were
An MHI is a real-time imaging technique that builds a map of carried out to show the correlation between biospeckle activity
movement based on the recency of motion in an image sequence. and quality indices or physiological processes. The first work in
This technique was proposed as an alternative to the routine this area was presented by Oulamara et al. [7], who showed
online methods of biospeckle image analysis [35]. In an MHI, pixel differences in temporal speckle evaluation among different com-
intensity is a function of the motion history at that location, where modities, i.e., an apple, an orange and a tomato. Xu et al. for the
brighter values correspond to more recent motion. The images first time showed that biospeckle could be used for monitoring the
stored in a buffer are first processed. The processing procedure age and shelf life of botanical specimens (oranges, potatoes,
consists of thresholding of the silhouette formed by the subtrac- apples, radishes, tomatoes) because the temporal fluctuations of
tion of two sequenced images. The final MHI image is created speckles decrease with aging [44]. In this experiment biospeckles
using the procedure of weighting the threshold images stored in were recorded in free-space propagation geometry, using a polar-
the buffer, with respect to the “lifetime” of each image. Tested in ized He–Ne laser as a source of coherent light and a CCD array
biological and non-biological samples, the MHI showed better camera interfaced to an image processor. The results were con-
results when compared to the LASCA. Additionally, the MHI firmed by Zhao et al. for an apple, a tomato and soybean sprouts
provided similar results to the already known offline methods and leafs [45]. Aging and decrease in vitality causes a decrease in
such as Fujii's and GD. biospeckle activity. This paper also reported that higher biospeckle
activity indicates higher flow rate in the veins of botanical
samples, thus this method can also be used for monitoring
2.2.5. Analyses in spectral domains nutrient transportation. Zhao et al. [45] recorded biospeckles by
Due to the complexity of biological materials, alternative the point-wise and whole-field methods using different system
techniques of biospeckle activity analysis, related on spectra arrangements. Both systems also used a low power He–Ne laser
domains have been developed [5,15,25,36]. Transformation of with a wavelength of 633 nm. In the case of the whole-field
the biospeckle signal to the frequency domain can be performed method, the laser beam was expanded and the speckle intensity
directly by the Fourier transform [36,37], or specifically by the was recorded as a function of time at three arbitrary pixels. Auto-
wavelet transform [5,15,25] which allows creation of new markers correlation coefficients, the cut-off frequency of the biospeckle
related to physical or biochemical phenomena. The transformation temporal spectrum and the contrast were calculated as the
to spectra domain could be performed in conjunction with tradi- parameters characterizing activity.
tional biospeckle laser methods like Fujii, GD and THSP [15]. This A typical shelf life experiment for apples showed a very
allows the spatial analysis of biospeckle activity with typical pronounced decrease in biospeckle activity with days of storage
indices, like IM, within chosen spectral bands. The spectral
[16] and that the activity correlates very well with the firmness of
characteristics of biospeckle activity showed promising ability to fruits [46]. Rabelo et al. have monitored the senescence process of
distinguishing of processes related with water activity of maize
an orange in the postharvest time and in this case also the
and beam seeds [15], separation of embryo and endosperm in biospeckle followed the biological variations [47]. Two applied
maize seeds [25] and observation, in frequency domain, aging of
approaches for quantification of the temporal biospeckle fluctua-
apple, pear and tomato [38]. tion (the inertia moment of the co-occurrence matrix and the
Using wavelet entropy as the descriptor of the dynamic speckle
statistical cumulant of the auto-correlation function of the spatial
phenomenon, as described in the section on global measures of temporal speckle pattern STS) showed a decrease in bioactivity
speckle activity, Passoni et al. showed a new method of evaluating
during orange storage. Moreover, tested fruits exhibited different
activity images [12]. The WE value was calculated over each time values of the measured speckle pattern dynamics, depending on
series that corresponded to each image pixel, allowing the
the illuminated point. Authors concluded that calculation of the
identification of different activity level regions. The different pixel activity obtained for the central calomel base (apex) allowed
gray levels of the resulting image are weighted according to the
differentiation between oranges according to their freshness.
WE values. The high activity regions (brighter areas) corresponded Obtained results suggested that biospeckle might be used in
to higher descriptor values, medium image values (middle gray
packing houses as an alternative quality evaluation tool.
pixels) indicate lower activity and the lowest descriptor values
(darkest regions) correspond to the scene's inert background. 3.2. Relation of biospeckle activity with biochemical processes

Aging and the shelf life of botanical samples are accompanied

3. Applications of the biospeckle-based methods in practical by various biochemical processes. Therefore, observed biospeckle
evaluation of agricultural crops activity could reflect the dynamics of metabolic-related changes in
tissue. However, interpretation of biospeckle activity from this
Since the biospeckle technique is non-invasive and nondes- point of view is very scarce. Few studies have shown that
tructive to biological tissue, it is considered as a valuable tool for biospeckle activity depends significantly on biochemical changes
monitoring various biological processes. It has been used for occurring during the postharvest storage of apples. Starch granule
example in medicine, most often for blood circulation measure- degradation caused a lesser number of scattering centers in the
ments [39–42] or in microbiology for bacterial (Pseudomonas apple tissue and resulted in smaller biospeckle activity, evaluated
aeruginosa) motility assessment [43]. For agricultural crops, which on the basis of cross-correlation calculation [17]. It has also been
are the topic of this review, various studies were carried out to proven that storage temperature affects measured biospeckle
show the usefulness of the method for monitoring aging (or shelf fluctuation [19]; calculated indices of apple biospeckle dynamics
life), detection of defects and diseases, evaluation of stages of plant (i.e., speckle contrast, moment of inertia and correlation coeffi-
development and in general to understand the phenomenon from cient) decreased when the sample was cooled down. In other
the biochemical point of view. words, freezing the tissue causes biospeckle activity to stop. In the

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

above studies similar experimental devices were used. A low rot infection and the third stage was when biospeckle activity
power red laser illuminated the apple surface and biospeckle suddenly decreased due to extensive rotting of the tissue. The
activity was recorded with a CCD camera with a strongly defo- increase in biospeckle activity due to infection was also observed
cused objective. The angle of incidence of the laser beam on the on apparently healthy parts of the fruit. Precise monitoring of the
botanical specimen was equal to 301 and the obtained average infected part of the apple tissue proved that early infection caused
speckle size was adjusted to make it much larger than the pixel an increase in biospeckle activity, which can be used for non-
size of the camera. Influence of chlorophyll content on measured destructive detection of apple diseases.
apple biospeckle activity was investigated by Zdunek and Herp- Biospeckles were also implemented for the detection of fungal
pich [18]. They stated that biospeckle activity linearly decreased contamination in seeds [36,49]. Braga Jr. et al. adopted the GD and
with an increase in chlorophyll content, which could be related to IM methods, as well as Fujii's algorithm, for evaluation of the
chlorophyll red light absorption. This effect limits the amount of biospeckle fluctuation in beans' seed [49]. Obtained results
light that can penetrate the tissue. This result confirmed the showed the capability of the employed technique for the identi-
previous results of Romero et al. [48] in tomato ripening degree fication of microorganisms' presence in beans. Seeds inoculated
determination. with fungi exhibited greater values of IM (higher biospeckle
activity) compared to the control group. In addition, two of the
3.3. Defects and diseases inspection three used fungi species could be distinguished by a distinct range
of IM values. The GD and Fujii's method image techniques also
The presence of surface and under surface defects alters displayed the presence of fungi. Rabelo et al. [36] analysed the
biological activity and can affect observed biospeckle variations. registered biospeckle images of bean seeds, artificially contami-
Pajuelo et al. tested damage in apples by using several dynamic nated by fungi, using an inertia moment method and the
speckle techniques [9]. They used quantitative methods, including frequency values of the STS signals, derived using a fast Fourier
autocorrelation of the temporal history of speckle patterns. The transform (FFT). The authors obtained similar results as Braga Jr.
moment of inertia of the co-occurrence matrix and the statistical et al. [49]; intensities of IM values of inoculated seeds were greater
cumulants calculation showed consistent results; after a mechan- than those of disease-free seeds. FFT coefficients also showed
ical injury biospeckle variations were lower than before and these differences between contaminated and healthy beans. However,
fluctuations decreased with time. To display the activity in a the authors concluded that IM values allowed the distinction
qualitative way and to visualize the damaged zones, the weighted between different fungi species, whereas frequency analysis was
generalized difference (WGD), the LASCA and the Konishi method useful in this matter only within certain harmonics.
were used. These techniques showed similar results; the bruised
region could be distinguished from the undamaged surrounding 3.4. Plant development
areas. Later Passoni et al. analysed bruising in apples by means of
entropy wavelet methods and they also obtained satisfactory Information on the application of biospeckle for monitoring
results, allowing for quantitative and qualitative estimation of plant development is scarce.
the sample bioactivity [12]. Briers suggested the use of laser speckle fluctuations as a
The possibility of biospeckle technique application for monitor- screening test in assessing the suitability of specimens for holo-
ing apple fungal infection development has been recently reported graphic measurement of growth or motion [50]. Braga et al.
by Adamiak et al. [20]. An experimental device was equipped with evaluated ability of this technique to detect changes in biological
a laser (8 mW, 635 nm) with an expanded beam, which illumi- activity across root tissues [24]. Roots of Coffea arabica and
nated apples at six places along the equator. A CCD camera Eucalyptus grandis plants were grown in gel tubes and were
recorded 4 s films to calculate the correlation coefficient. 100 illuminated by an expanded laser beam (10 mW He–Ne laser,
apples were monitored during cold storage for a long period with 632 nm) using a forward and backscattering configuration. Images
special attention to the development of bull's eye rot. Three stages were collected by CCD cameras and analysed by generalized
of biospeckle activity were observed (Fig. 6): the first one was difference GD, Fujii and LASCA methods. These approaches were
when biospeckle activity decreased due to shelf life-like behavior assessed visually by their ability to discriminate between different
(as described previously), the second stage was when biospeckle regions of the roots. The most satisfactory maps of biological
activity increased significantly due to first symptoms of bull's eye activity in root tissues were obtained using the GD method. The
results showed that biospeckle could be a quantitative indicator of
root cells molecular activity.
Recently, Szymanska-Chargot et al. monitored apple develop-
ment with the use of a biospeckle technique [21]. Biospeckle
activity increased during pre-harvest apple development. A sig-
nificant correlation was obtained between measured biospeckle
activity and soluble solid content, starch content and firmness,
showing that this method has the potential to be used for non-
destructive evaluation of these properties in the pre-harvest
period. The device for dynamic speckle pattern acquisition con-
tained an expanded diode laser beam (8 mW, λ¼ 635 nm) and a
CCD camera to record AVI films of biospeckle fluctuation. Bios-
peckle activity (BA) was evaluated based on a correlation coeffi-
cient calculation.

4. Biological bases of the biospeckle phenomenon

Fig. 6. Bull's eye rot development monitored with biospeckle activity BA during
cold storage of apples for 140 days. This is schematic profile based on monitoring
100 apples at 6 spots on the equator. The significance of the changes was confirmed As was summarized by Aizu and Asakura [51], specific char-
by ANOVA. acteristics of biospeckles are: multiple scattering of light in living

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
8 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

tissue, different dynamics of speckles observed on one specimen size of these objects according to scattering theory is optimal as a
and dependence of dynamics on the structure and activity of scattering centers and corresponds to the wavelength of laser light
objects, which may carry useful information about the biological used [52]. However, biospeckle could also arise from the scattering
and physiological activity of living objects. Statistical analysis of of light on other organelles, macromolecules and inorganic particles
the biospeckle phenomenon can therefore be applied not only to with sizes in the scale of nano- and micrometers. This observation is
non-contact measuring of specific characteristics of living speci- supported by the studies of Zdunek and Herppich [18] and Cybulska
mens but also, regarding the sensitivity of organisms to environ- and Zdunek [17], who showed a correlation between starch and
mental conditions, in the bioindication of the environment. chlorophyll content in apples and biospeckle activity. An increase in
According to the type of tested organism, the processes under- biospeckle activity with a decrease of chlorophyll content was
lying the phenomenon may be different. In the case of small caused by a decrease in red laser light absorption, deeper penetra-
moving organisms and independent, individual cells moving tion of the apple tissue by light, multiplication of internal reflections
passively (e.g., blood cells), biospeckle dynamics is caused mainly on moving cellular elements and finally, an increase in the
due to light scattering on the moving objects, while the dynamics dynamics of the interference image. In turn, an increase of bios-
observed in stable cell colonies or tissues is mainly the result of peckle dynamics with an increase in starch content increases the
cellular transport, organelles movement and Brownian motion and number of moving scattering centers in the tissue and in conse-
diffusion inside the objects [7]. In practice, a biospeckle image is a quence biospeckle activity. Further examples of the application of
mixture of speckles with different fluctuations of intensity biospeckle in the evaluation of fruit and vegetable quality have been
depending on its origin. As a fast component of biospeckle presented in a number of publications [16,46,48,65,66]. The exis-
dynamics, random physical movements are considered, while a tence of a relationship between biospeckle activity and qualitative
slow component could be caused mainly by the movement of a parameters in fruits and vegetables (which are, in fact, a group of
biological origin. The biological source of biospeckles was con- physical and chemical properties of plant tissues dependent on
firmed in experiments showing their dependence on temperature plant metabolism) during storage or ripening also suggest a
[19], the wavelength of the light used, the color of the object [52] biological origin of this phenomenon.
and light intensity [7]. According to the assumption that biospeckle is the result of
Taking into account attempts at the practical application of movement inside living organisms and that this movement has a
biospeckle, the most numerous and most advanced studies have mainly biological character (excluding Brownian motion and
been conducted in the medical field. Applications in the field of diffusion), several of the most likely, extra- and intracellular
medicine, comprehensively described in [53], include blood flow sources of light scattering, dependent on plant metabolism can
velocity measurements in different organs [42,54–56] as well as be identified.
muscular activity [57]. Moreover, biospeckle technique has been Long distance water transport velocity in vascular bundles can
proven to be a valuable tool for cancer tissue detection [25,58]. reach values up to few millimeters per second [67]. Peak flow rates
Less numerous but more diverse experiments have been carried in xylem are about 1mm/s, although maximum velocities as high
out in other areas of biology and agriculture, which relate as 0.8m/s have been reported [68]. Therefore, water transport in
primarily to the two groups of research: measurements of motility xylem vessels can cause vibration in the conductive elements and
of microorganisms and the evaluation of different properties of surrounding tissues because of cavitation [69–71], which might be
plants. The main principle of the application of biospeckle in the source of biospeckle dynamics.
determining microorganism motility is measuring the dynamics of Another potential process, which could be the basis of bios-
interference pattern formed by the scattering of light on the peckle, is cytoplasmic streaming. This directed flow of cytosol and
moving objects. The faster the microorganism motility, the higher organelles can reach even up to 100 mm/s [72]. Organelles, as well
biospeckle activity is. This method was used in the detection as dissolved macromolecules moving with the flow, are scattering
of fungi [49] and the viability of gametes [59], crustaceans centers, which can cause dynamic light scattering. Cytoplasmic
[60], parasites [61] and microorganisms in liquid culture media streaming and other processes connected with intercellular move-
[62,63]. ment occur as a result of the functioning of motor proteins.
Interpretation of biospeckle dynamics in the case of their A mechanism based on actin and myosin is mainly responsible
application to plants is much more difficult. Activity of interfer- for the formation of cytoplasmic streaming and division of the cell
ence pattern observed on plants is an effect of the changes of while microtubules with miosins and kinesins provides vesicle
optical properties of cell compartments, physical and biological transport and cell spatial reorganization. Motor proteins are there-
mechanisms connected with movement inside the objects and fore potentially the most important, but not the only source of
changes in plant structure due to growth or turgor regulation. biospeckle in cellular scale.
Optical properties, like the number of scattering and absorbing Ion channels and proton pumps also should be included as
elements and the high variability of densities in intracellular potential cellular elements that may affect biospeckle activity.
compartments, affect the path of photons and change the number Regulation of cell turgor and changes in the volume of specific
of potential interactions between light and matter, while the intracellular compartments is a direct result of channel and ion
movements of tissue elements of physical and biological origin, pumps functioning and can also result in dynamic light scattering.
increase the dynamics of the biospeckle image. Thus, the bios- Extracellular processes that are closely related to the metabo-
peckle can carry information about both, the structure and lism of cells connected with cell wall growth and degradation can
biological activity of the object. However, this information is be source of biospeckle as well. Excretion of cellulolytic enzymes
encoded in the complex spatio-temporal changes of the interfer- relaxing the cell wall and increase of turgor pressure during
ence pattern. growth, according to acid growth theory [73,74] can cause the
Experiments with the use of biospeckle as a tool for studying displacement of cell wall material, which can be the direct source
the structure of plants, refers mainly to the measurement of of dynamic light scattering. On the other side, chemical changes
growth and movement in plant organs [51,64]. However, most accompanying cell wall degradation during aging or damage,
researchers attempt to apply biospeckle in monitoring the proper- leading to changes in the spectral properties of the tissue in the
ties of plants connected with internal physiological processes. direction of greater light absorption, could decrease the amount of
In early work by Briers, one can find information that biospeckle diffusively reflected light and in consequence decrease biospeckle
dynamics arise mainly due to chloroplast movements, since the dynamics.

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 9

On the basis of the presented theoretical information one can easy to achieve connection to prevent the effects of vibrations.
conclude that in organs or parts of organs with high metabolism In addition, new cheaper system based on optical mouse are
rate e.g. during growth or turgor regulation, biospeckle activity is introduced [76].
relatively high, while in organs with a high pigment content or in Both, laboratory and industry applications require very accurate
the final stages of development biospeckle activity is relatively interpretation of biospeckle phenomena. Former studies consid-
low. This observation allows for a theoretical indication of the ered this issue in a very general way, i.e., they interpreted
potential practical applications of biospeckle. In the case of fruits biospeckle activity as cyclosis and Brownian motion movement
or parts of plants consisting mostly of parenchyma, biospeckle affected by pigment absorption. Biological samples are very vari-
activity probably reflects the general rate the sum of metabolic able and undergo very pronounced pre- and postharvest changes,
processes, which allows for an overall assessment of the physio- for example starch and chlorophyll degradation or changes in
logical state of these objects and identification of areas with water content and surface roughness. Therefore, the question
different biospeckle activities as a result of damage or pathogen which quality-related processes are responsible for the change in
attack. However, when the part of the plant consists of a variety of biospeckle activity is very difficult. To find the answer there are
tissues, it may also be possible to assess the differences in activity two ways: first, by working on model or simplified systems and
between the tissues in which the sources of biospeckle are second to improve the frequency analysis of biospeckle activity to
different or the rate of the processes in tissues is variable. An deconvolute signals for distinct biological processes.
example of a potential application of biospeckle is the study of leaf In summary, the biospeckle method shows plenty of interesting
physiology, where, completely non-destructively, changes in the nondestructive applications in agricultural crops. At present, in the
rate of water transport in vascular bundles or changes of turgor in area of quality evaluation, this method is still under development
stomatal cells or pulvini should be visible. and has a chance to be commercially used as it is already utilized
However, the use of biospeckle entails a number of problems, in medicine. However, industry scale application for agricultural
among which the most important are: determination of the crops requires more specific studies to obtain calibration or
quantitative relationship between the rate of a biological process classifications models dedicated to a certain commodity.
and biospeckle activity and the distinction between levels of
activity for individual biological processes and the sum of all the
processes that are sources of dynamic light scattering in the object References
[75]. The solution to both problems may be the development of
new measurement methods with the use of frequency analysis [1] Draijer M, Hondebrink E, van Leeuwen T, Steenbergen W. Review of laser
speckle contrast techniques for visualizing tissue perfusion. Laser Med Sci
and basic research concerning the relationship between specific
2009;24:639–51.
biological processes and biospeckle activity. [2] Briers JD. A note on the statistics of laser speckle patterns added to coherent
and incoherent uniform background fields, and a possible application for the
case of incoherent addition. Opt Quant Electron 1975;7(5):422–4.
[3] Briers JD. Laser speckle contrast imaging for measuring blood flow. Opt Appl
5. Conclusions 2007;37(1–2):139–52.
[4] Briers JD. The statistics of fluctuating speckle patterns produced by a mixture
Several reports about the use of biospeckle technique in the of moving and stationary scatterers. Opt Quant Electron 1978;10(4):364–6.
[5] Braga Jr. RA, Horgan GW, Enes AM, Miron D, Rabelo GF, Barreto Filho JB.
analysis of agricultural crops are presented in this review. Experi- Biological feature isolation in biospeckle laser images. Comput Electron Agr
ments showed that biospeckle activity correlates with physiologi- 2007;58:123–32.
cal changes during development and ripening and with defects [6] Rabal H, Arizaga R, Cap N, Grumel E, Trivi M. Numerical model for dynamic
speckle: an approach using the movement of the scatterers. J Opt A-Pure Appl
and diseases. From this point of view one can state that the Opt 2003;5:381–5.
method has important potential to be used as a nondestructive [7] Oulamara A, Tribillon G, Doubernoy J. Biological activity measurements on
method of quality control or as an indicator of maturity state. botanical specimen surfaces using a temporal decorrelation effect of laser
speckle. J Mod Optic 1989;36(2):165–79.
However, an important obstacle for agriculture industry level [8] Arizaga R, Trivi M, Rabal H. Speckle time evolution characterization by the co-
application is a lack of good calibration or classification models. occurrence matrix analysis. Opt Laser Technol 1999;31:163–9.
The reason for this is two-fold. First, the origin of biospeckle [9] Pajuelo M, Baldwin G, Rabal H, Cap N, Arizaga R, Trivi M. Bio-speckle
assessment of bruising in fruits. Opt Laser Eng 2003;40:13–24.
activity is a very complicated interaction between light and a
[10] Braga RA, Nobre CMB, Costa AG, Sáfadi T, da Costa FM. Evaluation of activity
botanical sample, where both elastic and non-elastic scattering through dynamic laser spackle using the absolute value of the differences. Opt
occurs, additionally affected by biochemical absorption. This is the Commun 2011;284:646–50.
case of course for other optical methods, however biospeckles [11] Braga RA, Silva BO, Rabelo G, Costa RM, Enes AM, Cap N, et al. Reliability of
biospeckle image analysis. Opt Laser Eng 2007;45:390–5.
have the additional factor of physical particle movement, which is [12] Passoni I, Dai Pra A, Rabal H, Trivi M, Arizaga R. Dynamic speckle processing
difficult to quantify for the preparation of calibration models. using wavelets based entropy. Opt Commun 2005;246:219–28.
Actually, the sensitivity for particle movement makes this method [13] Nobre CMB, Braga Jr. RA, Costa AG, Cardoso RR, da Silva WS, Sáfadi
T. Biospeckle laser spectral analysis under inertia moment, entropy and
particularly suitable for monitoring processes related, for example, cross-spectrum methods. Opt Commun 2009;282:2236–42.
with water movement in vessels, which occurs relatively faster [14] Braga RA, Silva WS, Sáfadi T, Nobre CMB. Time history speckle pattern under
than processes like pigment or polysaccharide degradation and statistical view. Opt Commun 2008;281:2443–8.
[15] Cardoso RR, Costa AG, Nobre CMB, Braga RA. Frequency signature of water
therefore is not affected by these biochemical changes. A second activity by biospeckle laser. Opt Commun 2011;284:2131–6.
problem is sensitivity to external vibrations. Most of the setup and [16] Zdunek A, Muravsky LI, Frankevych L, Konstankiewicz K. New nondestructive
methods of data analysis need,many images to evaluate biospeckle method based on spatial-temporal speckle correlation technique for evalua-
tion of apples quality during shelf-life. Int Agrophys 2007;21:305–10.
activity. When some vibration in the system occurs a large [17] Zdunek A, Cybulska J. Relation of biospeckle activity with quality attributes of
decorrelation is detected. apples. Sensors 2011;11:6317–27.
The most important advantage of this method is its nondes- [18] Zdunek A, Herppich WB. Relation of biospeckle activity with chlorophyll
content in apples. Postharvest Biol Technol 2012;64:58–63.
tructive character and the fact that it involves an uncomplicated
[19] Kurenda A, Adamiak A, Zdunek A. Temperature effect on apple biospeckle
and flexible system. A low power laser (a few mW), which very activity evaluated with different indices. Postharvest Biol Technol
often is expanded from a diameter of  1 mm to several centi- 2012;67:118–23.
metres, cannot usually heat large samples such as fruit, for [20] Adamiak A, Zdunek A, Kurenda A, Rutkowski K. Application of the biospeckle
method for monitoring bull's eye rot development and quality changes of
example. The configuration laser-object-detector can be setup apples subjected to various storage methods-preliminary studies. Sensors
horizontally or vertically and its simplicity makes it relatively 2012;12:3215–27.

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i
10 A. Zdunek et al. / Optics and Lasers in Engineering ∎ (∎∎∎∎) ∎∎∎–∎∎∎

[21] Szymanska-Chargot M, Adamiak A, Zdunek A. Pre-harvest monitoring of apple [49] Braga Jr. RA, Rabelo GF, Granato LR, Santos EF, Machado JC, Arizaga R, et al.
fruits' development with the use of the biospeckle method. Sci Hortic Detection of fungi in beans by the laser biospeckle technique. Biosyst Eng
2012;145:23–8. 2005;91:465–9.
[22] Fujii H, Asakura T, Nohira K, Shintomi Y, Ohura T. Blood flow observed by [50] Briers JD. Speckle fluctuations as a screening test in the holographic measure-
time-varying laser speckle. Opt Lett 1985;10(3):104–6. ment of plant motion and growth. J Exp Bot 1978;29(2):395–9.
[23] Fujii H, Nohira K, Yamamoto Y, Ikawa H, Ohura T. Evaluation of blood flow by [51] Aizu Y, Asakura T. Bio-speckle. In: Consortini A, editor. Trends in Optics.
laser speckle image sensing. Appl Optics 1987;26(24):5321–5. Research, Development and Applications. San Diego: Academic Press; 1996.
[24] Braga RA, Dupuy L, Pasqual M, Cardoso RR. Live biospeckle laser imaging of p. 27–49.
root tissues. Eur Biophys J 2009;38(5):679–86. [52] Briers JD. Wavelength dependence of intensity fluctuations in laser speckle
[25] Braga RA, Cardoso RR, Bezerra PS, Wouters F, Sampaio GR, Varaschin MS. patterns from biological specimens. Opt Commun 1975;13:324–6.
Biospeckle numerical values over spectral image maps of activity. Opt [53] R.A. Braga Jr., G.F. Rabelo, J.B.B. Filho, F.M. Borém, J. Pereira, M. Muramatsu and
Commun 2012;285:553–61.
I.M. Dal Fabbro, Applications in Biological Samples, In: H.J. Rabal and R.A.
[26] Arizaga R, Cap N, Rabal H, Trivi M. Display of the local activity using dynamical
Braga Jr., (Eds), Dynamic Laser Speckle and Applications, 2009, Boca Raton:
speckle patterns. Opt Eng 2002;41:287.
CRC Press, 181-220.
[27] Saúde AV, Menezes de FS, Freitas PLS, Rabelo GF, Braga RA. On Generalized
[54] Aizu Y, Asakura T. Bio-speckle phenomena and their application to the
Differences for Biospeckle Image Analysis. In: Proceedings of the 23rd
conference on graphics, patterns and images (SIBGRAPI), Gramado, Brazil; evaluation of blood flow,. Opt Laser Technol 1991;23:205–19.
2010. p. 209–15. [55] Briers JD. Laser Doppler, speckle and related techniques for blood perfusion
[28] Martí-López L, Cabrera H, Martínez-Celorio RA, González-Peña R. Temporal mapping and imaging. Physiol Meas 2001;22:35–66.
difference method for processing dynamic speckle patterns. Opt Comm [56] Murari K, Li N, Rege A, Jia X, All A, Thakor N. Contrast-enhanced imaging of
2010;283:4972–7. cerebral vasculature with laser speckle. Appl Optics 2007;46:5340–6.
[29] Briers JD, Webster S. Laser speckle contrast analysis (LASCA): a non-scanning, [57] Tanin LT, Rubanov AS, Markhvida IV, Dick SC, RachkovskyLI. In: von Bally G, S.
full-field technique for monitoring capillary blood flow. J Biomed Opt 1996;1 Khanna (Editors), Optics in Medicine, Biology and Environmental Research. In:
(2):174–9. Proceedings of the International conference on optics within life sciences
[30] Cheng H, Yan Y, Duong TQ. Temporal statistics analysis of laser speckle images (OWLS I), Elsevier Science Publishers B. V. Amsterdam; 1993. p.149.
and its application to retinal blood-flow imaging. Opt Express 2008;16 [58] Lee TK, Tchvialeva L,; Zeng H, McLean DI, Towards automatic detection of
(14):10214–9. malignant melanoma by laser speckle.In: Proceedings of the SPIE 7387,
[31] Zakharov P, Völker AC, Wyss MT, Haiss F, Calcinaghi N, Zunzunegui C, et al. Speckle 2010: Optical Metrology, 73871D (September 13). 2010; doi:10.1117/
Dynamic laser speckle imaging of cerebral blood flow. Opt Express 2009;17 12.870692; http://dx.doi.org/10.1117/12.870692.
(16):13904–17. [59] Carvalho PHA, Barreto JB, Braga Jr RA, Rabelo GF. Motility parameters
[32] Ramírez-San-Juan JC, Huang YC, Salazar-Hermenegildo N, Ramos-García R, assessment of bovine frozen semen by biospeckle laser (BSL) system. Biosyst
Muñoz-Lopez J, Choi B. Integration of image exposure time into a modified Eng 2009;102:31–5.
laser speckle imaging method. Phys Med Biol 2010;55(22):6857–66. [60] Ebersberger J, Weigelt G, Li Y. Coherent motility measurements of biological
[33] Dunn AK, Bolay H, Moskowitz MA, Boas DA. Dynamic imaging of cerebral objects in a large volume. Opt Commun 1986;58(2):89–91.
blood flow using laser speckle. J Cereb Blood Flow Metab 2001;21(3):195–201. [61] Pomarico JA, Di Rocco HO, Alvarez L, Lanusse C, Mottier L, Saumell C, et al.
[34] Cheng H, Luo Q, Zeng S, Chen S, Cen J, Gong H. Modified laser speckle imaging Speckle interferometry applied to pharmacodynamic studies: evaluation of
method with improved spatial resolution. J Biomed Opt 2003;8:559–64.
parasite motility. Eur Biophys J 2004;33(8):694–9.
[35] Godinho RP, Silva MM, Nozela JR, Braga RA. Online biospeckle assessment
[62] Zheng B, Pleass C, Ih C. Feature information extraction from dynamic
without loss of definition and resolution by motion history image. Opt Laser
biospeckle. Appl Optics 1994;33:231–7.
Eng 2012;50:366–72.
[63] Cole JA, Tinker MH. Laser speckle spectroscopy—a new method for using
[36] Rabelo GF, Enes AM, Braga Jr. RA, Dal Fabro IM. Frequency response of
small swimming organisms as biomonitors. Bioimaging 1996;4:243–53.
biospeckle laser images of bean seeds contaminated by fungi. Biosyst Eng
[64] Briers JD. The measurement of plant elongation rates by means of holographic
2011;110:297–301.
[37] Sendra GH, Arizaga R, Rabal H, Trivi M. Decomposition of biospeckle images in interferometry: possibilities and limitations. J Exp Bot 1977;28(2):493–506.
temporary spectral bands. Opt Lett 2005;30(13):1641–3. [65] Bergkvist A, Biospeckle-based Study of the Line Profile of Light Scattered in
[38] Ansari MZ, Nirala AK. Assessment of bio-activity using the methods of inertia Strawberries. Master degree thesis, Faculty of technology at lund University,
moment and absolute value of the differences. Optik 2013;124:512–6. Sweden, 1997.
[39] Stern MD. In vivo evaluation of microcirculation by coherent light scattering. [66] Braga Jr RA, DalFabbro IM, Borem FM, Rabelo G, Arizaga R, Rabal HJ, et al.
Nature 1975;254:50–8. Assessment of seed viability by laser speckle techniques. Biosyst Eng 2003;86:
[40] Asakura T, Takai N. Dynamic laser speckle and their application to velocity 287–294.
measurement of diffuse object. J Appl Phys 1981;25:179. [67] Taiz L, Zeiger E. Plant Physiology. 4th ed. Sunderland: Sinauer Associates; 2006
[41] Briers JD, Fercher AF. Retinal blood-flow visualization by means of laser [Chapter 4].
speckle photography. Invest Ophthalmol Vis Sci 1982;22:255–9. [68] Nobel PS. Physicochemical and Environmental Plant Physiology. San Diego:
[42] Zhu D, Lu W, Weng Y, Cui H, Luo Q. Monitoring thermal-induced changes in Academic Press; 483–96.
tumor blood flow and microvessels with laser speckle contrast imaging. Appl [69] Milburn JA, Johnson RPC. The conduction of sap. II. Detection of vibrations
Optics 2007;46:1911–7. produced by sap cavitation in Ricinus xylem. Planta 1966;69:43–52.
[43] Sendra H, Murialdo S, Passoni L. Dynamic laser speckle to detect motile [70] Holbrook NM, Ahrens ET, Burns MJ, Zwieniecki MA. In vivo observation of
bacterial response of Pseudomonas aeruginosa. J. Phys: Conf Ser cavitation and embolism repair using magnetic resonance imaging. Plant
2007;90:012064. Physiol. 2001;126:27–31.
[44] Xu Z, Joenathan C, Khorana BM. Temporal and spatial properties of the time– [71] Davis SD, Sperry JS, Hacke UG. The relationship between xylem conduit
varying speckles of botanical specimens. Opt Eng 1995;34(5):1487–502. diameter and cavitation caused by freezing. Am J Bot 1999;86(10):1367–72.
[45] Zhao Y, Wang J, Wu X, Williams FW, Schmidt RJ. Point-wise and whole-field [72] Shimmen T, Yokota E. Cytoplasmic streaming in plants. Curr Opin Cell Biol
laser speckle intensity fluctuation measurements applied to botanical speci- 2004;16:68–72.
mens. Opt Laser Eng 1997;28:443–56. [73] Rayle DL, Cleland RE. Enhancement of wall loosening and elongation by acid
[46] Zdunek A, Frankevych L, Konstankiewicz K, Ranachowski Z. Comparison of solutions. Plant Physiol 1970;46:250–3.
puncture test, acoustic emission and spatial-temporal speckle correlation [74] Rayle DL, Cleland RE. The Acid Growth Theory of Auxin-induced Cell Elonga-
technique as methods for apple quality evaluation. Acta Agrophysica
tion Is Alive and Well. Plant Physiol 1992;99:1271–4.
2008;11:303–15.
[75] Kurenda A, Pieczywek PM, Adamiak A, Zdunek A, Effect of cytochalasin B,
[47] Rabelo GF, Braga Jr. RA, Fabbro IMD, Arizaga R, Rabal HJ, Trivi MR. Laser
lantrunculin B, colchicine, cycloheximid, dimethyl sulfoxide and ion channel
speckle techniques in quality evaluation of orange fruits. Rev Bras Eng Agríc
inhibitors on biospeckle activity in apple tissue, Food Biophys 2013; http://dx.
Ambient 2005;9:570–5.
[48] Romero GG, Martinez CC, Alanis EE, Salazar GA, Broglia VG, Alvarez L. Bio- doi.org/10.1007/s11483-013-9302-7.
speckle activity applied to the assessment of tomato fruit ripening. Biosyst Eng [76] da Silva MM, De R, Nozela Almeida, Chaves MJ, Braga Jr. RA, Rabal HJ. Optical
2009;103:116–9. mouse acting as biospeckle sensor. Opt Commun 2011;284(7):1798–802.

Please cite this article as: Zdunek A, et al. The biospeckle method for the investigation of agricultural crops: A review. Opt Laser Eng
(2013), http://dx.doi.org/10.1016/j.optlaseng.2013.06.017i