Efficient Detection and Classification of PEST Using Image Thresholding and Edge Detection Technique

In these modern days many techniques have been developed to save agricultural fields, but the present scenario is mainly focusing on organic food products for a healthier life, so it is a challenging task to detect the pests for cultivating organic crops as well as inorganic crops. In this paper two approaches have been developed for detecting the pest, one approach is using thresholding and edge detection through image processing techniques, by utilizing thresholding the presence of pest is identified. It gives the range of pixels for the clear identification of the image. In our work we have considered an average value of 100 pixels. Using the technique of edge Detection, the outline of the pest is determined. The other approach is convolution neural network (CNN) which provides pest detection by three steps, firstly direct wavelet transform (DWT) next neural network detection last is area detection. By using these techniques pest detection is developed.