Air Pollution Detection

Isaac Njuguna Kimaru Patricia Kasuku

Technical University of Mombasa Technical University of Mombasa
[email protected] [email protected]

Abstract- Pollution from industries, vehicles, change of of the main contributors of hazardous air pollutants.
weather, indoor pollution among other sources Other pollutants include particulate matter, chlorine,
negatively impacts the health of employees and sulphur dioxides, hydrogen sulphides, carbon
neighbouring communities and the potential for adverse
dioxide, carbon monoxides and nitrous oxides and
health outcomes is heightened when the industries are
located in rural areas where the bulk of the population
hydrogen sulphide.
is vulnerable because of limited information about their A significant number of roads and industries in this
rights and limited capacity to defend themselves or
country are in close proximities to heavily populated
influence policy decisions. Furthermore, the situation is
compounded by the fact that developing countries have human settlements and residential areas. A
limited air quality management systems due to substantial percentage of the local population is
inadequate legislation and lack of political will, among therefore continuously exposed to polluted air, as a
other challenges. Maintaining a balance between result of industrial activities and vehicular emissions.
economic development and sustainable environment is a Continuous exposures to hazardous air pollutants
challenge, hence investments in pollution prevention often result in various forms of ailments including
technologies get side lined in favour of short-term respiratory diseases. Ailments usually result in loss of
benefits from increased production and job creation.
productivity and sometimes loss of life. Air
This lack of air quality management capability
translates into lack of air pollution data, hence the false
pollutants can further be transported over long
belief that there is no problem (Omanga et al). This distances and frequently end up in terrestrial and
however, is not the case and only further conceals a aquatic ecosystems, through wet and dry deposition.
major public health crisis Sulphur dioxide, nitrogen dioxide and particulate
matter are some of the prevalent air pollutants in
The purpose of this project is to examine the use of
Kenya, especially in urban areas and industrial sites.
digital cameras to detect the magnitude of atmospheric
pollution present in the atmosphere. Digital cameras are
They have been widely associated with adverse
inexpensive and are being used in countless areas, many effects on human health and the environment
of which are outdoors and very public.The theory including, aggravation of asthma and chronic
presented in this paper could help agencies to monitor bronchitis respiratory ailments, eye and throat
air quality at any of these sites.The project aims at irritation, acidification of aquatic ecosystems through
availing information to policy makers and researchers acid rain and reduction of soil fertility, acceleration
concerning public awareness and attitudes towards of corrosion and destruction of some components of
environmental pollution pertinent to development and buildings, inflammation of lung tissues, pulmonary
implementation of environmental policies for public
fibrosis, respiratory irritation, impairment of lung
health.
defences, respiratory diseases and eye irritation.
Index Terms –charge coupled device (CCD),
complementary metal oxide semiconductor (CMOS), Astoundingly, most developing countries, especially
Rayleigh scattering, Mie scattering in sub-Saharan Africa, do not have air quality
management systems because of inadequate
I. INTRODUCTION legislation, budgetary constraints and lack of political
Air pollution is one of the major and most prevalent will, among other things (Mulaku GC, 2001). In these
forms of environmental pollution worldwide regions, there is a major challenge in maintaining a
including in Kenya. Industrial activities and balance between economic development and a
emissions, vehicles, construction and agricultural sustainable environment; hence investments in
activities, incineration and windblown dust are some pollution prevention technologies like emission

1
controls are commonly outweighed by the short-term industrial facilities and environmental health
benefits that accrue from increased production and perception indicators. Questions were developed
job creation. The lack of air quality management based on variables representing aspects of perceived
capabilities in these regions translate into lack of air health risks identified from the above literature
pollution data, which in most cases gives the false pertinent to the study area and population.
belief that it is not a problem. (Eunice Omanga,
2014) This however, is not the case and only further
conceals a major public health crisis in the Load
developing world. Continuou the
sly capture images
There are many researches done in environmental Start sky images into
analysis and how pollutants affect the atmosphere
Matlab
around us. However very few give an easy and
inexpensive way by which one can analyse the air
quality we breathe, hence this project proposes an
easy and cheap method of pollution detection using
Matlab software and digital cameras.
Display
The theory is based on how certain molecules and the
particles that are present in clean air absorb, loaded
luminesce, refract, reflect, or scatter the red, green,
image
and blue (RGB) visible light spectrum in a
measurable manner. The longer wavelength
components (red side) of visible light through the
atmosphere are scattered less than the shorter Split into RGB
wavelength components (blue side). The blue channels
component is scattered more than the other colour
components (and, thus, is responsible for blue sky).
The longer wavelength components of visible light
are also refracted less than the shorter wavelength
components. The presence of certain pollutants and Get histogram
suspended particles in air will cause different levels values for each
of absorption, re-emission, refraction, or scattering in channel
the RGB spectrum than that for cleaner air.
II. METHODOLOGY

A digital camera was used for the purpose of data Plot values
collection. Data was collected in form of raw images
in one plot
taken in the selected county of Mombasa. The sites
are; Bamburi, Nyali and Tudor. There after the
images were analysed and compared against images
taken Nyali, assumed to free of any pollution.
The quantitative survey instrument was a
Stop
questionnaire developed primarily by reviewing and
adapting questionnaires from literature. For
environmental perception, literature reviewed
included previous risk perception, environmental
health and environmental psychometric surveys
within and outside Kenya which were then modified
for relevance. The survey questionnaire consisted of
a list of items reported in literature as known or
potential environmental health hazards from
2
 considerable risk to them despite the fact that the
economy of the area largely depended on the factory.
III. RESULT Respondents also argued that they had not been
By analysing the shift between red, green and blue actively involved in identifying solutions to the
histograms of polluted sky and unpolluted sky. The environmental challenges.
peak in every histogram represents the average of Table 3.1: Selected demographic characteristics of
red, green and blue components in an image. the study population
Compare by how many times the peak has been
shifted for a polluted sky in comparison with the
unpolluted sky. Thefigure below shows the histogram
of unpolluted sky image (image1) and polluted sky
image (image2 ). It can be clearly observed that the
histograms of red, green and blue are overlapping
with each other in the case of a polluted sky. On the
other hand, histograms of red, green and blue of the
unpolluted sky from Nyali shows non-overlapping
histograms. This shows that pollutants which are
mixed in the air are causing the red, green and blue
components values to be changed.

Figure 1.1 : Pure image RGB histogram

Figure 1.2: Polluted image RGB histogram

IV. CONCLUSION
This is because pollutants present in the atmosphere
cause the red and green components to scatter and The presence of certain pollutants and suspended
attenuate the blue component, thus absorbing the blue
particles in air will cause different levels of
component and promoting the red component. Thus
blue component histogram shifts to the left and red absorption, re-emission, refraction, or scattering in
histogram shifts to the right in the case of polluted
the RGB spectrum than that for cleaner air. With the
sky. The more shift, the more polluted the sky.
From the analysis of the questionnaire, ver 80% of help of image processing techniques, pollution can be
respondents perceived industrial pollution as posing a
detected in the atmosphere using methods which are
3
useful to analyse hidden information in an image.
These methods include-
Analysing the RGB colour channel histogram of an
image
Analysing the shift in channels in comparison with
an unpolluted sky image.

REFERENCES
A. Prochazka, M. K. J. F. P. H. K. H., 2000. Satelite Image
Processing and Air Pollution Detection. Istanbul, IEEE.
Amritphale, A. N., 2013. A Digital Image Processing Method for
Detecting Pollution in the Atmosphere from Camera Video.
UNLV Theses, Dissertations, Professional Papers, and
Capstones. 1793.
Anita kulkarni, T. R. T., 2014. Automated System for Air
Pollution Detection and Control in Vehicles. International
Journal of Advanced Research in Electrical, Electronics and
Instrumentation Engineering, III(9).
C. H. LIU, A. J. C. &. G. R. L., 1996. An image-based retrieval
algorithm of aerosol characteristics and surface reflectance for
satellite images. International Journal of Remote Sensing,
17(17), pp. 3477-3500.
C.J. Wong, M. M. K. A. a. H. L., 2009. Development of Low Cost
Air Quality Monitoring System by Using Image Processing
Technique. In: Y. Chen, ed. Image Processing. Croatia: Intech,
pp. 106-114.
Eunice Omanga, L. U. Z. B. M. G., 2014. Industrial air pollution in
rural Kenya: community awareness, risk perception and
associations between risk variables. BMC Public Health , pp.
2-3.
Harada, K. H. T. a. K., 2007. Haar Wavelet Based Approach for
Image Compression and Quality Assessment of Compressed
Image. IAENG International Journal of A
pplied Mathematics.
Mulaku GC, K. L., 2001. Mapping and Analysis of air Pollution in
Nairobi. Nairobi, Institution of Surveyors of Kenya.
Sіnecen, M., 2016. Digital Image Processing with MATLAB. In:
Image Processing . Aydin,: Intech.
Teng., L. M., 2011. The Environmental Impacts of Construction
Activities.
Wandiga, P. S. O., 2015. Air Quality Assessment in Selected Sites
in Kenya. Nairobi, s.n.
WHO Regional Office for Europe, 2013. Review of evidence on
health aspects of air pollution –REVIHAAP Project,
Copenhagen: World Health Organization