Abstract: During the production of mandarin cans， some membrane\|removed mandarin segments have defects such as membrane residues， seed residues or segment damage， which would affect the quality of products. Currently， factories adopt manual inspection to reject defective segments， which increased the cost and may cause hygiene problems. Based on LabVIEW， a machine vision program for the automatic sorting of membrane\|removed mandarin segments was designed. An experimental platform was built with color GigE camera， white ring light and red flat light， to test inspection accuracy and duration. It was shown that the accuracy of inspection program reached （99.1±0.5）%， and the average time consumption was （17±5） ms. The inspect speed was able to achieve （64±19） pieces per second.

Key words: LabVIEW, machine vision, automatic sorting algorithm, membrane\, removed mandarin segments