Development of on-line apple bruise detection system

Abstract

This research aims to develop an on-line apple bruise detection system, using chlorophyll fluorescence images for nondestructive apple bruise detection and analysis. The on-line apple bruise detection system is consisted of a fluorescent image acquisition unit, a fruit carrying revolving tray unit, a conveying unit, and a control unit. The LabVIEW graphical programming language was used to build the control program, and Vision Assistant was used for image processing and analysis. The wavelength of the fluorescence excitation source used was 460 nm. A color camera captured the fluorescence released from Golden Delicious apple, and the wavelength was 680 ± 10 nm. The fruit was revolved by the fruit carrying revolving tray, 60° each time, revolved six times, so as to shoot the complete apple surface. The image processing procedures of bruise recognition include image preprocessing, medium filtering, regional binarization, background removal, noise filtering and hole filling for recognition, and analysis of bruise blocks. This study applied 67.9 ± 4.84N, 82.1 ± 4.57N and 102 ± 8.44N impact forces to apples respectively to make bruises. The test results showed that the proposed on-line apple bruise detection system could recognize the apple bruise 100% after 45 min from the impact on apples. The prototype system could inspect 92 apples for bruise automatically hourly on an average. This study has successfully built an on-line apple bruise detection system, and the inspection of the system may be accelerated in the future for commercial uses.