0% found this document useful (0 votes)
101 views9 pages

PCB Checking 3

Uploaded by

Anurdha shammika
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
101 views9 pages

PCB Checking 3

Uploaded by

Anurdha shammika
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Accelerat ing t he world's research.

Automatic O ptical Inspection f or


PCB M anufacturing : a Survey Eid M.
Taha , E. Emary , Khalid Moustafa
Eid Taha

Related papers Download a PDF Pack of t he best relat ed papers 

An aut omat ic opt ical inspect ion syst em for t he diagnosis of print ed circuit s based on neural …
Alessandra Lera

Aut omat ic opt ical inspect ion for det ect ing defect s on print ed circuit board inner layers
Aoy AuTomat ize

COMPUT ER MACHINE VISION INSPECT ION ON PRINT ED CIRCUIT BOARDS FLUX DEFECT S
Med Fridhi
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1095
ISSN 2229-5518

Automatic Optical Inspection for PCB


Manufacturing: a Survey
Eid M. Taha, E. Emary, Khalid Moustafa

Abstract— As the rapid development in electronic industries based on Printed Circuit Board (PCB) designs and high volumes manufacturing
capacities and the need for high quality products with minimum defect rate comes the importance of Automated Optical Inspection (AOI)
technology. The basic objectives among different AOI system manufacturers are to improve lighting, computing capability, flexibility of part staging
and vision software. These improvements make AOI products more intelligent, flexible, and with far more repeatable results that are superior to
human visual inspection. For finding of errors in PCB many algorithms are proposed in literature. A variety of approaches for automated visual
inspection of printed circuit have been reported over the last three decades. The last reported survey in this topic introduced by Moganti96 that is
why the need to introduce this survey to cover reported work after 1996. Also Moganti survey covers solely bare PCBs visual inspection.
In this survey, algorithms and techniques for the automated inspection of printed circuit boards are examined. A classification of these algorithms is
presented and the algorithms are grouped according to this classification. This survey concentrates mainly on image analysis and fault detection
techniques; these also include state-of-the-art algorithms.

Index Terms— Automatic Optical Inspection, Printed Circuit Boards, Visual Inspection, Fault Detection, Image Analysis, Surface Mount Technology

1 INTRODUCTION Mount Technology (SMT) has been widely used in the


electronic industry recently. As a result, the electronic

A utomated Optical Inspection (AOI) inspects a wide


range of products, such as Printed Circuit Boards
components rely on the solder joint to provide the electrical
connection to the PCBs; therefore, the quality of the solder

IJSER
(PCBs), Liquid Crystal Displays (LCDs), transistors, joint can be critical to the quality of the electronic
automotive parts, lids and labels on product packages or components [1]. Automatic Optical Inspection (AOI) of
agricultural products (seed corn or fruits). In case of PCB- solder joints has been a critical issue for quality control in
inspection, a camera autonomously scans the device under PCB assembly as AOI has the enormous potential of
test (DUT) for variety of surface feature defects such as completely automating human visual inspection procedures
scratches, stains, open circuits, short circuits and thinning [2, 3].
of the solder as well as missing components, incorrect
components, and incorrectly placed components. In general, PCB inspection can be divided into three
categories: reference comparison approach, design rule
A machine vision or an AOI system can acquire millions of checking (non-referential) approach, and hybrid approach
data points (pixels) in a fraction of a second. These data [6].
points are used for visual inspection and precision
measurement. The reference comparison approach is based on a
comparison between the image of the PCB to be tested and
AOI visually scans the surface of the PCB. The board is lit that of an ideal PCB which is conform to pre-defined design
by several light sources and observed by a scanner or by a specifications. While it is a fast method for comparison and
number of high definition cameras. This enables the doesn't require any CAD data but any miss-orientation of a
monitoring of all areas of the board, even those hidden in board’s placement may lead to wrong analysis in reference-
one direction by other components. It should be noted that based method [5]. There are two major techniques: image
each manufacturer of AOI systems utilizes different comparison methods and model-based inspection. Image
inspection algorithms and lighting techniques, each of these comparison, which is the simplest approach, consists of
systems may have varying strengths and weaknesses comparing both images pixel-by-pixel using simple logic
depending upon the item/product it inspect. operators such as XOR. The main difficulty found in these
techniques is determining a precise alignment of the
Assembly of Printed Circuit Boards (PCBs) using Surface reference image and the test image, which makes its
utilization difficult. More sophisticated proposals under the
———————————————— same idea, involve feature and template matching [23], but
• Eid M. Taha, Information Technology Dept, Faculty of Computers & suffer from the same problem and normally require a large

• E. Emary, Information Technology Dept, Faculty of Computers &


Information, Cairo University, Egypt, E-mail: [email protected] number of templates. Model-based methods are techniques,
which match the pattern under inspection with a set of
• Khalid Moustafa, Information Technology Dept, Faculty of Computers &
Information, Cairo University, Egypt, E-mail: [email protected]
predefined models. They are also called Graph-Matching
Information, Cairo University, Egypt, E-mail: [email protected] Methods [23] and are based on the structural, topological,
and geometrical properties of the image. The major difficulty
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1096
ISSN 2229-5518

of those methods is related to the matching complexity. 2 Shorts/bridges


Although Sun and Tsai [Sun and Tsai, 1992] proposed a 3 Missing conductor
technique called Pattern Attributed Hyper graph to make the 4 Missing Component
method more practical, it still remains a complex and time- 5 Component with wrong polarities or
consuming method. flipped components
6 Wrong Component Placed
The design rule checking approach is based on the 7 Incorrect hole dimension
verification of the general design rules that is essential in the 8 Missing hole
verification of the widths of conductors and insulators. As a POTENTIAL 9 Partial Open 9.1 Mouse bit
kind of automatic inspection algorithm for bare PCB, the 9.2 Nicks
design rule checking has been proposed and well known to
9.3 Pinholes
the automatic visual inspection system manufacturers [25].
10 Excessive 10.1 Specks
The design rule checking (DRC) method checks if all patterns
spurious 10.2
and spaces of PCB surface meet or violate common
Spurs/protrusions
knowledge, which is called the design rule. Because a simple
10.3 Smears
algorithm is applied directly to an image, the
11 Pad 11.1 Under etching
implementation of this algorithm is comparatively easy. This
violations 11.2 Over etching
means that it does not require severe alignment and
adjustment of a mechanical part to get a non-distorted image 11.3 Breakout
and can be easily parallelized. However, this method is a 12 Variations 12.1 Small thickness
very time consuming process and a great computing power between the wiring
is needed to meet user’s requirement of inspection time. printed lines 12.2 Large conductors
12.3 Excessive
Hybrid approach merges the advantages of the reference conductors
comparison method and the DRC method to overcome the 12.4 Incipient short
(conductor too close)

IJSER
weaknesses of each method. For example, most of the
design-rule verification methods are limited to verifying Dividing the inspection types according to the fed data type

• Automatic visual/optical inspection: based on


minimum conductor trace, angular errors, and spurious entered into the inspection system as follows:
copper. PCB defects which do not violate the design rules are
detected by reference comparison methods. These methods visual system (camera scanning) this system can
can detect missing features or extraneous features. The detect many types of PCB defects for both bare PCBs
design rule process detects most type of defects within small and PCBs with SMT components surface related like
and medium features while the comparison methods are (breaks, missing components, open, short,
sensitive to the largest features. Hybrid approach makes use component with wrong polarities, over etching,
both of these methods as they complement each other and under etching). Visual inspection may require
therefore achieve a high sensitivity of PCB inspection. special installation for cameras and lighting

• Automated X-ray inspection: this type of inspection


Another way of classification is Front End and Back End [36] conditions.
processing. When a computer vision system is proposed for
the automatic detection, localization, and segmentation of uses X-ray as a scanning source for the PCB and find

• ICT (In-Circuit Test) based inspection: Is an example


solder joints on PCBs under different illumination un-visual PCB defects related to ICs and BGAs
conditions. This is referred to as the “front-end” inspection
system. The back-end involves the classification of the solder of white box testing where an electrical probe tests a
joints. The front-end inspection system includes illumination populated printed circuit board (PCB), checking for
normalization which is applied to an image to effectively and shorts, opens, resistance, and capacitance.
efficiently eliminate the effect of uneven illumination while
keeping the properties of the processed image the same as in
2 PCB INSPECTION METHODS:
the corresponding image under normal lighting condition.
Consequently special lighting and instrumental setup can be In this part we will summarize most of recent work on
reduced in order to detect solder joints. different PCB inspection approaches. Figure 1 explores the
most common PCB inspection techniques into three
Table 1 explores most PCB error both for Bare PCBs [35] and categories reference comparison approach, design rule
PCBs with SMT components. checking (non-referential) approach, and hybrid approach
[6].
TABLE 1, PCB ERRORS CLASSIFICATION
As described in introduction Referential techniques divided
into two types Image comparison and Model based. For non-
FATAL 1 Breaks 1.1 Fracture
Referential has we recorded three types. Encoding
1.2 Cut
Techniques include (Run-length, Boundary analysis),
1.3 Scratches
Morphological process based and Hierarchical Marked Point
1.4 Cracks
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1097
ISSN 2229-5518

Process. Finally the Hybrid techniques, we summarize sum 53sec (in Sun Sparc5 with 32 MB of memory) even with
of their techniques (Generic, Genetic based, Learning proposal of no (reading and writing files) and expected time
methods, circular pattern matching, Log gabor filter and about 5 sec. also many types of defects can't be covered
Classifier Fusion, Pattern detection using boundary analysis). (mouse bite, under etching, over etching, Variations between
Below we will describe techniques not included in the printed lines).
Moganti96 Survey.
Zuwairie and Syed [40] proposed an automated visual
printed circuit board (PCB) inspection system based on a
two-dimensional HAAR wavelet transform. During the
inspection, two type of images are required; reference and
tested PCB images. The wavelet transform is applied to the
reference and tested PCB images. According to previously
proposed wavelet-based PCB inspection algorithm, a
reference comparison between the reference and tested PCB
images has been done in wavelet domain by employing
image difference operation. However, in practice, this
operation also bring along the unwanted noise due to
misalignment and uneven binarization. Thus, in this system,
for the real-time implementation, the image difference
operation between the reference and tested PCB images is
replaced with image subtraction. The output of the image
subtraction operation can be differentiated as positive,
negative, and zero images. By doing so, a noise elimination
procedure is applied to positive and negative images,
resulting noise free positive and negative images. Lastly, the
noise free images are subjected to XOR logic operation to

IJSER
generate the output of the inspection system. This system is
well suited for small and medium scale PCB manufacturers
where the sophisticated alignment facilities are hard to
purchase.

Noor and Zuwairie [6] proposed an algorithm to group the


defects found on bare PCB. Using a synthetically generated
PCB image, the algorithm is able to group 14 commonly
Fig. 1 classification of PCB inspection techniques.
known PCB defects into five groups. The proposed algorithm
includes several image processing operations such as image
2.1 REFERENTIAL METHODS: subtraction, image adding, logical XOR and NOT, and flood
Reference based techniques based on point or feature fill operator. The algorithm needs to be improved to achieve
comparison between a reference image (golden image) and better result, which is to classify all the 14 types of defects
input image. It requires much time for image processing and separately. Also unwanted noise should be considered.
enhancement; specially image registration task before image
comparison applied. Fabiana and Flávio [7] present a Computer Vision system for
printed circuit board (PCB) automated inspection in the
Mauro and Roberto [38] introduced a technique for PCB proposed Computer Vision PCB Inspection System they
inspection based on the comparison of the connected table (is considered the first manufacturing stage, i.e., the board
a list of a list of connected PCB holes) of a Reference Image printing in PCB manufacturing process. They first compared
and a Test Image. The method is based on connected a PCB standard image with a PCB image, using a simple
component analysis, which is a natural way to extract the subtraction algorithm that can highlight the main problem-
connectivity information of the conductors of a PCB. The regions. Then they used connection analysis to find fatal and
registration of the PCB holes, which is a common problem potential errors, like breaks and circuit shorts. In other to
related to referential model techniques, is solved by the develop this methodology in real PCB, they propose to
concept of zone of influence of each hole (zone of influence of magnify the problem-regions and start to find the errors in a
a point is the region where all the points are nearer to that set of PCB sections, which are smaller than the main PCB
point than any other point in the image) and Labeling to image. This approach seems to be very effective if applied in
extract the conductor connectivity of the PCB image. The a real time inspection system. Therefore, they propose a new
implementation of this technique is using standard algorithm to solve PCB inspection problem, considering its
Morphology Image Processing techniques. This technique is efficiency in reducing the computational time. It is fast
impractical for manufacturing as it has long processing time algorithm but constraint only on PCB printed process while
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1098
ISSN 2229-5518

most of PCB defects happen in the subsequent processes.


Also they confirm that the environment lighting should be
uniform and that all inspected PCB belong to the same
category. And will be necessary to calibrate the system every
time we change the reference PCB or environment
illumination.

Lin and Zhou[42] proposed an algorithm for PCB inspection


based on Mathematical Morphology. In this work, the
conception of linkage information table was presented which
is regarded as correlation features of template PCBs. This
work also includes pre-processing and feature extractor.
Compared with features between detected images and
template images based on linkage information table, those
defects existed in tested PCBs can be identified, which has
achieved the automatic identification and real time
inspection. Proved by some experiments, this algorithm will
largely compress feature data volume, decrease correlation
workload and decrease data processing time. Through
experiments, this algorithm can inspect ordinary defects fast
and precisely. At last, this work has these following
conclusions:
1) The noises are diminished in all images by using
median filtering method, which has following
merits: small data number, short analyzing time,
simple algorithm.

IJSER
2) The defect inspecting method based on linkage
information table of comparison with template and
tested images was presented, which need images
not to be located at each other precisely, and largely Fig. 2 Block diagram of the PCB inspection system
decrease the required precision on hardware.
After images have been captured, image registration is
Ismail and Syed [16] proposed an algorithm for detecting six executed to obtain a well-aligned defective image. Then,
types of defects (pin hole, missing hole, short circuit, under- image subtraction operation, minimum thresholding and
etch, open circuit and mouse-bite) on bare PCBs based on median filtering are used to get noise-free positive and
Image subtraction technique. Figure 2 represent the detection negative images. Next, defect classification algorithms that
steps: have been taken from (Ismail and Zuwairie [43]) are
employed to classify six printing defects as mentioned. The
system was proven to be an alternative way to efficiently
detect and classify defects. Furthermore, the system is cheap
as mechanical alignment facility is no need to purchase.

2.2 NON-REFERENTIAL METHODS:


The non-reference inspection methods do not need any
reference image or pattern to work with. These methods are
also called the design-rule inspection technique. The patterns
and tracks of inspected image are checked with the design
specification standards. Non-reference inspection methods
can avoid the disadvantages of reference methods, but these
methods may miss defects that do not violate the design
rules.
In 1996, Borba and Facon proposed a non-reference
inspection method. It can detect defects without considering
a reference board. The research succeeded in verifying
vertically, horizontally and 45 degrees oriented traces.
Csaba [13] introduced a probabilistic approach for optical
quality checking of solder pastes (SP) in Printed Circuit
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1099
ISSN 2229-5518

Boards (PCB) based on Hierarchical Marked Point Process Fourier descriptors to establish correspondences between the
(HMPP). Dealing with unregistered image inputs, it is target objects detected in the reference and the inspected
required to address at the same time SP identification, and images. To enhance the computational efficiency, the
detection of special soldering errors, called scooping. proposed component detection and contour tracing
Scooping occurs when the squeegee blade enters into the algorithms use the run length encoding (RLE) and Blobs
stencil aperture and removes paste from the center of the tables to represent the pixel information in the regions of
print. This is caused by high print pressure and can be interest. The Fourier descriptors derived from the component
exaggerated by rubber blades and large stencil apertures. A boundaries are used to match the target objects. Finally, the
single scoop in a PCB does not cause a critical quality transformation parameters for aligning the inspected image
problem, however, if the number and summarized volume of with the reference image are estimated based on a novel
such artifacts surpass a given threshold, the board should be phase-shifted technique. Experimental results show that the
withdrawn. For this reason he introduced a novel proposed FDBIA algorithm sustains similar accuracy as
Hierarchical Marked Point Process (HMPP) framework, achieved by the commercial software Easyfind against
which is able to handle the paste and scooping extraction various rotation and translation conditions. Also, the
problems simultaneously, so that the SPs and included computational time consumed by the FDBIA algorithm is
scoops have a parent–child relationship. A global significantly shorter than that by Easyfind.
optimization process attempts to find the optimal
configuration of entities, considering the observed data, prior Wu and Chen [29] proposed an Image-array-based automatic
knowledge, and interactions between the neighboring circuit optical inspection system captures multiple images of the
elements. The proposed method is evaluated on a real PCB inspected printed circuit board through synchronous
image set containing more than 3000 SPs and 600 scooping exposure. Design error of the mechanism always leads to
artifacts. A morphology-based baseline method is also geometric errors in the image such as translation, rotation
introduced for the problem and used as reference for and scaling distortion. These errors make severe
qualitative and quantitative comparison against the displacements in the image. To solve such problems, a
proposed model. Experiments reported in this work concise and rapid error correction algorithm is proposed base
on multiple point interpolation. Experiments show that the

IJSER
confirmed the superiority of the proposed HMPP model and
its usability for forthcoming industrial inspection systems algorithm can assure rapid and correct mosaic of images of
over tested Morph system. the whole printed circuit board.

2.3 HYBRID METHODS: Cai and Huang [30] proposed an effective framework for
PCB Automatic Optical Inspection. First, a simple strategy
Hybrid inspection methods involve a combination both of that establishes accurate registrations between template
reference and non-reference inspection methods. These image and inspecting images is proposed. Then they describe
methods have the advantages of these two techniques, but how to construct multi-layers PCB template based on cluster
hybrid inspection methods are too complex to implement. analysis which ensure that the color information is not lost.
Finally they discuss the existence and classification of defects.
Chen and Zhang [39] presented a smart machine vision Experimental results indicate that the registration speed and
(SMV) system for printed circuit board (PCB) inspection. It detection ability is greatly enhanced.
consists of two modules, LIF (Learning Inspection Features)
Chia-Te and Wen-Hao [32] researched for a new automated
and OLI (On-Line Inspection). The LIF module automatically
optical inspection system (AOI) to improve the inspection
learns inspection features from the CAD files of a PCB board.
rate of defects, which has multi-color illuminator, and can
The OLI module runs on-line to inspect PCB boards using a
capture images of the tested High Density Interconnection
high-resolution 2-D sensor and the knowledge provided by
the LIF components. Some example of learned features: Printed Circuit Board (HDI PCB) by using machine vision,
and also can identify several of defects quickly and
• Number of pads on each array, accurately by a new-designed image processing software
• Location of each pad, systems. The results of experiments prove that PCB defects
• Shape of each pad, inspection rate of the new AOI is improved and the false
• Fill or no-fill status of each pad, and
alarm rate of defects is reduced to 0.3%.
• Location of every fiducial points on each PCB image.
Syamsiah and Jonathan [5] presented a novel integrated
The system has been tested over 8000 boards on a system in which a number of image processing algorithm are
manufacturing line and the detection accuracy was above
embedded within a Genetic Algorithm (GA) based
97%.
framework in order to provide an adaptation and better
Chin and Chun [8] present a Fourier descriptor based image quality analysis with less computational complexity while
alignment algorithm (FDBIA) for applications of automatic maintaining flexibility to a broad range of defects. Their
optical inspection (AOI) performed in real-time environment. simulations on real PCB images demonstrate that the HGA is
It deliberates component detection and contour tracing robust enough to detect any missing components and cut
algorithms and uses the magnitude and phase information of
solder joint with any size and shape with significant
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1100
ISSN 2229-5518

reduction in computational time compared to conventional phase only. While Back-End includes the next step in visual
approaches. While it use a tailored hybrid GA (HGA) is used inspection process; classification and Recognition phase.
to estimate geometric transformation and needs classification
process to recognize the type of defect for decision making 3 FRONT-END AND BACK-END
by the system itself or human operator. CLASSIFICATION BASED:
Crispin and Rankov [41] investigate methodologies for S.Yarlagadda and C.Fookes [36] proposed two inspection
locating and identifying multiple objects in images used for modules for an automatic solder joint classification system.
surface mount device inspection. Their work focuses on The ‘‘front-end’’ inspection system includes illumination
automated object-recognition techniques for locating normalization, localization and segmentation. The ‘‘back-
multiple objects using grey-model fitting for producing a end’’ inspection involves the classification of solder joints
generalized template for a set of components. The work uses using the Log-Gabor filter and classifier fusion. Five different
the normalized cross correlation (NCC) template-matching levels of solder quality with respect to the amount of solder
approach and examines a method for constraining the search paste have been defined. The Log-Gabor filter has been
space to reduce computational calculations. The search for demonstrated to achieve high recognition rates and is
template positions has been performed exhaustively and by resistant to misalignment. This proposed system does not
using a genetic algorithm. need any special illumination system, and the images are
acquired by an ordinary digital camera. This system could
Wu Hao, Zhang Xianmin [4] proposed a neural network contribute to the development of automated non-contact,
combined with genetic algorithm for the diagnosis of solder non-destructive and low cost solder joint quality inspection
joint defects on printed circuit boards (PCBs) assembled in systems. Experimental results demonstrate that the proposed
surface mounting technology (SMT). Six types of solder joint system improves performance and robustness in terms of
have been classified in respect to the reality in the classification rates. However, further improvements can be
manufacture. The images of solder joint under test are expected from the current technologies. More robust features
acquired and 14 features are extracted as input features for could be developed by focusing on the errors of the current
the classification. The neural network is easily become over- algorithm and the use of multiple views of solder joints could

IJSER
fitting because these input features are not independent of also improve classification performance.
each other, so the genetic algorithm is introduced to select
and remove redundant input features. The experimental CONCLUSION:
results have proved that the neural network combined with
genetic algorithm reduced the number of input feature and As the rapid technology improvement of electronic
had a satisfying recognition rate. While it has a significant industry and heavy mass production the need for AOI
time for classification as a requirement for manufacturing. system arise as a need of quality and speed. In this survey
we tried to cover most recent researches and algorithms in
AOI system.

• A remarkable work appeared regarding of


detection Bare PCB and less for PCB with SMT
components defects detection. But still there are


some issues regarding speed and false alarms.
A few recent researches introduced regarding non-


referential techniques.
Most of the techniques require special environment


conditions.
A little research introduced regarding defects
classification.

REFERENCES
[1] Lee ZS, Lo RC (2002) “Application of vision image cooperated with
multi-light sources to recognition of solder joints of PCB”, The
2002 Conference on Technologies and Applications of Artificial
Fig. 3 Classification of PCB inspection techniques. Intelligence (TAAI 2002). pp. 425–430.

[2] Yang FC, Kuo CH, Wing JJ, Yang CK (2004) “Reconstructing the 3D
Figure 3 represents another view of visual inspection solder paste surface model using image processing and artificial
neural network”, IEEE international conference on Systems, Man
techniques based on [36]; Front-End and Back-End. Front-
and Cybernetics (10-13 Oct. 2004) ,The Hague, The Netherlands;
End examines techniques for image enhancement and Vol 3: pp. 3051-3056.
segmentation; that is those techniques related to detection
IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1101
ISSN 2229-5518

[3] Lin SC, ChouCH, Su CH (2007) “A development of visual flow-based motion analysis” Springer Journal Machine Vision
inspection system for surface mounted devices on printed circuit and Applications Vol 22 Issue 4, July 2011pp. 629-649.
board” Industrial Electronics Society, 2007. IECON 2007. 33rd
Annual Conference of the IEEE; Taipei Taiwan ; pp. 2440–2445. [16] Ismail Ibrahim, Syed Abdul Rahman, Syed Abu Bakar, Musa
Mohd Mokji, Jameel Abdulla Ahmed Mukred, Zulkifli md Yusof,
[4] Wu Hao,Zhang Xianmin, Kuang Yongcong, Ouyang Gaofei, Xie Zuwairie Ibrahim, Kamal Khalil, Mohd Saberi Mohamad (2012)
Hongwei (2013) “Solder joint inspection based on neural network “A printed Circuit Board Inspection System with Defect
combined with genetic algorithm” Elsevier, International Journal Classification Capability” International Journal of Innovative
for Light and Electron Optic 2013; Vol 124, Issue 20, October 2013, Management, Information & Production ISME International
pp. 4110–4116. March 2012, Vol 3, No 1: pp.82-87.

[5] Syamsiah Mashohor, Jonathan R. Evans and Ahmet T. Erdogan [17] Sharat Chandra Bhardwaj (2012) “Machine Vision algorithm for
(2006) “Automatic Hybrid Genetic Algorithm Based Printed PCB Parameters Inspection” National Conference on Future
Circuit Board Inspection”, IEEE, 2006. AHS 2006. First Aspects of Artificial intelligence in Industrial Automation
NASA/ESA Conference on Adaptive Hardware and Systems (NCFAAIIA 2012), Proceedings published by International
(AHS'06 Istanbul) 15-18 June 2006, pp. 390-400. Journal of Computer Applications® (IJCA),No 2, pp. 20-24

[6] Noor Khalid, Zuwairie Ibrahim, And Mohamad Abidin (2008) “An [18] Chern-Sheng Lin, Guo-Hao Loh, Chuen-Lin Tien, Tsung-Chih
Algorithm to Group Defects on Printed Circuit Board for Lin, Yih-Chih Chiou (2013) “Automatic optical inspection system
Automated Visual Inspection”. International Journal of for the micro-lens of optical connector with fuzzy ratio analysis”
Simulation Systems, Science & Technology IJSSST, Vol.9, No.2, Elsevier International Journal for Light and Electron Optic 2013
May 2008 pp. 1-10. Vol 124, Issue 17, September 2013, pp. 3085–3090

[7] Fabiana R. Leta, Flávio F. Feliciano, Flavius P. R. Martins (2008) [19] Chuan-Yu Chang, Senior Member, IEEE, Chun-Hsi Li, Student
“Computer Vision System for Printed Circuit Board Inspection” Member, IEEE, Si-Yan Lin, and MuDer Jeng, Fellow, IEEE (2009)
ABCM Symposium Series in Mechatronics - Vol. 3: pp. 623-632. “Application of Two Hopfield Neural Networks for Automatic
Four-Element LED Inspection” IEEE Transactions on Systems,
[8] Chin-Sheng Chen a, Chun-Wei Yeh a, Peng-Yeng Yin (2009) “A Man, And Cybernetics—Part C: Applications And Reviews, Vol
novel Fourier Descriptor Based Image Alignment Algorithm for 39, No 3, pp. 352-365, May 2009.

IJSER
Automatic Optical Inspection” Elsevier Journal of Visual
Communication and Image Representation Vol 20, Issue 3, April [20] Su-Hwan Kim a, Jee-Hyun Kim b, Shin-Won Kang (2011)
2009, pp. 178–189. “Nondestructive defect inspection for LCDs using optical
coherence tomography” Elsevier Displays Vol, Issue 5, December
[9] Chern-Sheng Lin, Su-ChiChang, Yun-LongLay, Mau-ShiunYeh, 2011, pp. 325–329.
Chi-ChinLin (2010) “A novel Positioning Method for Optical
Automatic Inspection of An LCD Assembly Process” Elsevier [21] C.W. Liao, Y.S. Tarng (2009) “On-line automatic optical inspection
International Journal for Light and Electron Optic (2010) Vol 121, system for coarse particle size distribution” Elsevier Powder
Issue 12, June 2010, pp. 1089–1095. Technology Vol 189, Issue 3, 13 February 2009, pp. 508-513

[10] Der-Baau Perng · Yen-Chung Chen (2010) “An advanced Auto- [22] Toshifumi Honda, Hisae Yamamura, Syouji Ueda,Minoru
inspection System for Micro-router Collapse” Springer Journal of Kawabata and Takanori Ninomiya (1994) “Automated Visual
Machine Vision and Applications Vol 21 Issue 6, October 2010 pp. Inspection System for Electronic Devices Mounted on Printed
811-824. Circuit Boards Using Light-Section Method”, MVA'94 IAPR
Workshop on Machine Vision Applications Dec. 13-15, 1994
[11] Guifang Duan · Yen-Wei Chen · Takeshi Sukegawa (2010) Kawasaki Japan. pp. 575-578
“Automatic optical flank wear measurement of microdrills using
level set for cutting plane segmentation” Springer Machine Vision [23] Moganti, M., Ercal, F., Dagli, C. H. and Shou, Tsunekawa (1996)
and Applications August 2010 , Vol 21 Issue 5: pp. 667–676. “Automatic PCB Inspection Algorithms: A Survey”, Computer
Vision and Image Understanding, Vol. 63, No.2, pp. 287 – 313.
[12] N.S.S. Mar, P.K.D.V.Yarlagadda, C.Fookes (2011) “Design and
development of automatic visual inspection system for PCB [24] Yung-Nien Sun, and Ching Tsorng Tsai (1992) “A New Model-
manufacturing” Elsevier Robotics and Computer-Integrated Based Approach for Industrial Visual Inspection”, Elsevier
Manufacturing Vol 27, Issue 5, October 2011, pp. 949–962. Journal of Pattern Recognition, Vol 25, Issue 11, November 1992,
pp. 1327–1336.
[13] Csaba Benedek (2011) “Detection of soldering defects in Printed
Circuit Boards with Hierarchical Marked Point Processes” [25] Ji-joong Hong, Kyung-ja Park and Kyung-gu Kim (1998) “Parallel
Elsevier Pattern Recognition Letters Vol 32, Issue 13, 1 October processing machine vision system for bare PCB inspection”,
2011, pp. 1535–1543. Industrial Electronics Society, IECON ’98. Proceedings of the 24th
Annual Conference of the IEEE Aachen Germany Vol 3, pp. 1346
[14] Rong-Fong Fung, Chih-Yu Yang, Ching-Tsung Lai (2011) “ – 1350.
Graphic supervisory control of an automatic optical inspection for
LED properties” Elsevier Measurement Vol 44, Issue 8, October [26] Lin, Chern-Sheng, Ho, Chen-Wei,Yang, Shih-Wei,Chen, Der-Chin,
2011, pp. 1349–1360. Yeh, Mau-Shiun (2010) “Automatic optical inspection system for
the image quality of micro lens array”, Indian Journal of Pure &
[15] Du-Ming Tsai, Hsin-Yang Tsai (2011) “Low-contrast surface Applied Physics. Vol 48, September 2010, pp. 635 – 643.
inspection of mura defects in liquid crystal displays using optical

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 1102
ISSN 2229-5518

[27] Bin Ren , Lianglun Cheng (2009) “SMT Automatic Optical IEA/AIE 2001 Budapest, Hungary, June 4–7, 2001 Proceedings,
Inspection Path Planning Based on MDSPSO Algorithm”, Vol 2070, pp. 513-518.
International Conference on Computational Intelligence and
Natural Computing 2009 Wuhan China Vol 2: pp. 134-137 [40] Zuwairie Ibrahim , Syed Abdul Rahman Al-Attas (2004)
“Wavelet-Based Printed Circuit Board Inspection System”,
[28] Yen-Chung Chen, Jiun-Hung Yu, Mu-ChiauXie and Fang-Jung International Journal of Signal Processing Vol 1 No 2, pp. 73-79.
Shiou (2011) “Automated optical inspection system for analogical
resistance type touch panel”, International Journal of the Physical [41] A. J. Crispin & V. Rankov (2007) “Automated inspection of PCB
Sciences Vol. 6(22), pp. 5141-5152, 2 October, 2011. components using a genetic algorithm template-matching
approach”, Springer International Journal of Advanced
[29] Jianjie Wu, Zhongbing Chen (2011) “A Concise and Rapid Error Manufacturing Technology (2007) Vol 35: pp. 293-300.
Correction Algorithm in Automatic Optical Inspection of Solder
Paste”, IEEE Proceeding ICICIS '11 Proceedings of the 2011 [42] Li-zong Lin, Luo-shan Zhou, Jian-ding Wan and Zhi-qin Qian
International Conference on Internet Computing and Information (2009) “Study of PCB Automatic Optical Inspection System Based
Services Hong Kong, Vol 1: pp. 325-329. on Mathematical Morphology”, IEEE International Conference on
Computer Technology and Development 2009, Kota Kinabalu
[30] Yu Cai, Yanjin Huang, Shugong Zhang (2012) “Research of Defect Malaysia, Vol 2:pp. 405-408.
Inspection and Processing in PCB Automatic Optical Inspection”,
Proceeding ICECC '12 Proceedings of the 2012 International [43] Ismail Ibrahim, Zuwairie Ibrahim, Kamal Khalil, Musa Mohd
Conference on Electronics, Communications and Control, Mokji, Syed Abdul Rahman Syed Abu Bakar, Norrima Mokhtar
Zhoushan China pp. 803-806. and Wan Khairunizam Wan Ahmad (2012) “An Improved Defect
Classification Algorithm For Six Printing Defects And Its
[31] Guangjie Xiong, Jingjing Lu, Siyuan Wu (2012) “New Automatic Implementation On Real Printed Circuit Board Images”
Optical Inspection System for PCB”, Proceeding ICECC '12 International Journal of Innovative Computing, Information and
Proceedings of the 2012 International Conference on Electronics, Control Vol 8, No 5(A), pp. 3239-3250, May 2012.
Communications and Control, Zhoushan China pp.1270-1273.

[32] Chia-Te Liao, Wen-Hao Lee, Shang-Hong Lai (2012) “A Flexible


PCB Inspection System Based on Statistical Learning”, Journal of

IJSER
Signal Processing Systems Vol 67 Issue 3, June 2012, pp. 279-290.

[33] Mika Liukkonen, Elina Havia, Yrjö Hiltunen (2012) “Review:


Computational intelligence in mass soldering of electronics - A
survey”, Journal Expert Systems with Applications: An
International Journal Vol 39 Issue 10, August, 2012, pp. 9928-9937.

[34] R. C. Gonzalez and Richard E. Woods (1993) “Digital image


processing” Book, Addison-Wesley Publishing Company Inc,
ISBN 0-201-50803-6.

[35] Ajay Pal Singh Chauhan, Sharat Chandra Bhardwaj (2011)


“Detection of Bare PCB Defects by Image Subtraction Method
using Machine Vision”, Proceedings of the World Congress on
Engineering 2011 Vol II pp1597-1601, WCE 2011, July 6 - 8, 2011,
London, U.K.

[36] N.S.S. Mar, C. Fookes, P.K.D.V. Yarlagadda (2009) “Design of


automatic vision-based inspection system for solder joint
segmentation”, Journal of Achievements in Materials and
Manufacturing Engineering, Vol 34 Issue 2, June 2009, pp. 145-
151.

[37] Hong Kyu Chung,Rae-Hong Park (1997) “Two-stage high-


precision visual inspection of surface mount devices”, Journal of
Electronic Imaging Vol 6(4), pp. 517–524 (October 1997).

[38] MAURO HIROMU TATIBANA, ROBERTO DE ALENCAR


LOTUFO (1997) “Novel Automatic PCB Inspection Technique
Based on Connectivity” IEEE, X Brazilian Symposium on
Computer Graphics and Image Processing, Campos do Jordao
Brazil, 14-17 Oct 1997, pp. 187 – 194

[39] Tie Qi Chen, JianXin Zhang, YouNing Zhou and Yi Lu Murphey


(2001) “A Smart Machine Vision System for PCB Inspection”, 14th
International Conference on Industrial and Engineering
Applications of Artificial Intelligence and Expert Systems,

IJSER © 2014
http://www.ijser.org

You might also like