Damaged Fruits Inspecting System by Machine Vision (Part 1)

Spectral Reflectance of Peach Defects

Abstract

This article describes the spectral reflectance from the surface of sound and defective peaches in order to develope an inspection system to select fruits by machine vision. Fruit-surface reflectances of peaches damaged by compression, puncture, friction, peeling and bruises were compared with sound ones, and then the prediction confidence intervals were confirmed. This made it clear which ranges of wave length should be adopted for the machine vision inspection system.
Measurement over the ranges of 800-940nm and 1140-1400nm infrared rays provided more than 5% differences in spectral reflectance between sound and four kinds of defective peaches, damaged by compression, puncture, peeling and bruises. These infrared regions were valid for detecting such defects.