Selection of color space for image segmentation in pest detection

Abstract

Image segmentation is the process of partitioning an image into multiple segments, so as to change the representation of an image into something that is more meaningful and easier to analyze. In modern crop status management in greenhouse, instead of doing manually, crop status is monitored using cameras with some automation. One of the major problems in the greenhouse crop production is the presence of pests. An accurate and timely monitoring of pests population is the basic requirement. In the pest detection, image analysis is very important and image segmentation is one of the desirable steps to distinguish the pest from rest of part of an image. Color image segmentation is desirable than gray scale image segmentation. This work suggests Entropy based thresholding in which the maximum information content is used to decide the segmentation rule. Results are dependent upon a color space selection. The suggested segmentation algorithm is applied for images of pest infected leaves.Results are compared with the results of Fuzzy c-mean method.