Key points are not available for this paper at this time.
ABSTRACT A laboratory machine vision system was developed to detect and identify surface defects (scar, cuts, bruise, scale, wormhole, and brown rot) on fresh market peaches. Image analysis algorithms were developed for segmenting defect regions in the peach images, and a classifier identified the segmented regions as specific defect types. Experimental tests were conducted to determine system accuracy in estimating defect area and identifying defect type. Sample correlation coefficients between predicted and manudly measured defect areas ranged from 0.56 for scale to 0.92 for brown rot. Classifier performance in identifying each segmented region as a member of one of eight classes (scar, stem cavity, cut, bruise, scale, wormhole, brown rot, and noise) was 31% error rate for the near-infrared system and 40% for the color system.
Miller et al. (Tue,) studied this question.