Study on color image processing based intelligent fruit sorting system

A machine vision system that can be used for automatic high-speed fruit sorting is proposed. Fruit area was first segmented out from image with an Ohta-color-space based thresholding algorithm; blob algorithm was utilized to remove noises in image; spline-interpolation based algorithm was adopted to detect fruit contour. In fruit sorting process, fruit´s color ratio, which was calculated with HSI color space, was selected as classification feature. Fruit sorting was realized by classic Bayes classifier, whose parameters were obtained by a study module. This system was tested with Crystal Fuji apples, and an average sorting accuracy of 90% was achieved.