A study on non-destructive method for detecting Toxin in pepper using Neural networks

Mycotoxin contamination in certain agricultural systems have been a serious concern for human and animal health. Mycotoxins are toxic substances produced mostly as secondary metabolites by fungi that grow on seeds and feed in the field, or in storage. The food-borne Mycotoxins likely to be of greatest significance for human health in tropical developing countries are Aflatoxins and Fumonisins. Chili pepper is also prone to Aflatoxin contamination during harvesting, production and storage periods. Various methods used for detection of Mycotoxins give accurate results, but they are slow, expensive and destructive. Destructive method is testing a material that degrades the sample under investigation. Whereas, non-destructive testing will, after testing, allow the part to be used for its intended purpose. Ultrasonic methods, Multispectral image processing methods, Terahertz methods, X-ray and Thermography have been very popular in nondestructive testing and characterization of materials and health monitoring. Image processing methods are used to improve the visual quality of the pictures and to extract useful information from them. In this proposed work, the chili pepper samples will be collected, and the X-ray, multispectral images of the samples will be processed using image processing methods. The term “Computational Intelligence” referred as simulation of human intelligence on computers. It is also called as “Artificial Intelligence” (AI) approach. The techniques used in AI approach are Neural network, Fuzzy logic and evolutionary computation. Finally, the computational intelligence method will be used in addition to image processing to provide best, high performance and accurate results for detecting the Mycotoxin level in the samples collected. This research paper gives an overview of the ongoing research in non-destructive methods for finding toxins in chili pepper by making a comparative study of the previous works.