Adaptive thresholding based segmentation of infected portion of pome fruit

Abstract

In this paper, we elaborate the segmentation of the infected portion of pome fruits. An adaptive thresholding method is applied on pears. Other segmentation methods such as OTSU, region growing, merge and split, greedy snake etc are available. But the main issue is the fixed threshold value for thresholding causes some of the information loss due to improper segmentation. A multi-level thresholding algorithm is developed in this paper. An improved thresholding algorithm not only reduces the processing time but, also improves the segmentation technique. The region of interest after segmentation is then further analyzed to identify the defected portion of pome fruit.