Classification of Fruit Images Using Optimal Features Extracted By Soft Computing Techniques

Abstract

newlineFruit classification is found to be one of the emerging areas in computer vision and newlineclassification. The accuracy of the classification system depends on the quality of acquired newlineimages, the number of features, types of features, selection of optimal features newlinefrom extracted features and type of classifier used. Images taken in poor environment newlineconditions decrease the visibility and hidden information details of digital images. newlineTherefore, image enhancement techniques are necessary for improving the significant newlinedetails of these images. This thesis provides a review to highlight the recent newlineprogress of image enhancement techniques for improving the visibility of fruit images. newlineIn this research work, a new Type-II Fuzzy based image enhancement technique is newlinedesigned to improve the quality of visibly weather degraded fruit images. The Type-II newlineFuzzy logic can automatically extract the local atmospheric light and roughly eliminate newlinethe atmospheric veil in local detail enhancement.