Development and Evaluation of a Small-Scale Apple Sorting Machine Equipped with a Smart Vision System

One of the most important matters in international trades for many local apple industries and auctions is accurate fruit quality classification. Defect recognition a key online computer-assisted sorting machines. Because cavity structure stem calyx regions, system tends to mistakenly treat them as true defects. Furthermore, there no small-scale machine with smart vision classification where it needed. Thus, current study focuses on highly feasible methodology based Niblack thresholding learning technique using k-nearest neighbor (k-NN) classifiers associated locally designed machine. To find an appropriate mode, effects different numbers k metric distances region detection were evaluated. Results showed effectiveness value Euclidean accuracy. It found that 5-nearest classifier distance 80 training samples produced best accuracy rates, at 100% 97.5% calyx. The significance result very promising fabricating advanced low-cost high horticultural industry.