Home
If you have any Suggestions/Comments please let me know, thanks.
gassoumi@yahoo.com
My research interests are in the areas of, Digital Signal Processing techniques including Image Processing, Pattern Recognition, Neural networks, Digital Communications, and Fuzzy Logic.
The purpose of my research work is to use soft computing approach to the develop an Integerated Pest Management System better known as IPM. This project is done in collaboration with Dr. John J. Ellington from the Department of Entomology, Plant Pathology, and Weed Science at New Mexico State University My advisor is Dr. Nadipuram R. Prasad.
The approach comprises a computer vision-based system integrated with soft
computing methods such as neural networks, fuzzy logic, and hybrid structures
for the indentification of insect classes. While the study has focused on
cotton fields, the approach can be extended to several other agricultural and
other image recognition applications as well.
The motive behind doing this research work is the wide spread use of pesticide
around the globe and the danger associated with such an application. A general
overview of the IPM system, and the hazadous associated with pesticide is laid
out in the following paragraphs.
What is an Integrated Pest Management (IMP)?
Integrated Pest Mangement better known IPM is a system that monitors pest
populations by using biological, cultural (crop rotating), and chemicals to
keep pests below economically damaging levels.
IPM enables farmers to monitor and control the pests in their fields, keeping
to an absolute minimum the use of expensive and potentially damaging and
and dangerous chemical pesiticides.
Why IPM?
Most pesticides kill beneficial insects as well as harmful insects which
suggested that they should be used only when necessary. Conserving beneficial
insects in cotton fields such as ladybeetles, greenlacewing, damsel bugs, and
parasitic is one component of integrated pest management.
The primary objective in IPM is to reduce the amount of pesticides by properly
monitoring of pest beneficial insect populations quickly. Such a method would
allow the evaluation and reliance on non-insecticide mortality factors for
biological insect control. The need for future pesticide applications would be
based on the insect population in the field.
Pesticides Hazardous and Dangers
Threat to human being
Pesticides are not only expensive and potentionally damaging to crops and the
environment, they can also be extremely dangerous for people-especially
farmers who may be unaware of of the toxicity of the chemicals they are using,
and who cannot afford the necessary protective clothing. Pesticides are
derived from nerve gases used in the second world war, and therefore, are
poisonous to humnan as well as to pests. The lag of knowledge regarding the
danger associated with pesticide can result in devastating results. In a recent
news (October 26, 1999), 20 children in Perou died because they consumed milk
that was stored in cans that was previously used to stock pesticide.
Threat to wildlife
Cotton fields and alfalfa provide food and cover for several species. Though
pesticides applied to these fields can improve cropy yields and quality, and
make farming profitable, it can potentially harm wildlife and their habitat.
Wildlife is directly at risk when they eat plants or seeds with chemicals
residues or when they swallow the pesticide granules themselves. They are
exposed indirectly when they eat insects or other animals killed by chemicals.
Wildlife that are in the field when the pesticides are appliled or, that enter
the field shortly afterward may inhale vapors or get pesticides on their skin
feathers, or in their eyes. Pesticide that are washed by rain into streams may
harm aquatic animals as well.
Animals can be exposed by swallowing or inhaling a chemical, or getting it on
the skin. Reducing pesticides use is therefore one of the best ways to protect
wildlife.
Water contamination
Pesticides enter surface and ground water primarily as runoff (The movement
of water over a sloping surface) or by leaching (the movement of pesticides
through the soil as opposed to movement over the surface). Runoff occurs when
water is applied to the soil at a faster rate than it can enter the soil.
Runoff water can carry pesticides in the water itself from crops by seeping
through the soil to ground water or entering streams as surface run off and are
most prevalent in agricultural areas. Without proper safegards, pesticides
have the potential to seriously threaten ground water supplies throughout the
cotton producing regions in the world. A vast majority of the world population
obtain drinking water from ground water sources and in excess of 90% of the
population in agricultural areas use ground water as their source of drinking
water.
Current Research in IPM Systems
Since the spacial patterns of insects is not fixed, it is difficult, in not
impossible to obtain satisfactory statistical models to predict or classify
insects categories. This also made more complex by changes in insect
population with season and time. Factors such as weather, host plants, soil,
predator, parasitoid, and behavioral factors of the insects population can all
contribute towards an extremely complex problem in insect classification.
Several attempts have been made in this area to come up with recognition
systems that would be capable of recognizing and classifying insects in the
cotton and alfalfa fields. Most of the early work was based on statistical
methods. While the results are satisfactory in classifying some insect
classes, results in other categories are marginal. Therefore, the there is
need for a model-free approach to insect classification.
In our work, the idea involves the use of a vision-based scheme in which the
sample of insects collected is identified into two groups, namely, those which
are classified as biological control species (good ones), and those which are
pests (the bad insects).