Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Environmental Microbial & Food Safety Laboratory » Research » Publications at this Location » Publication #304280
Title: A simple multispectral imaging algorithm for detection of defects on red delicious apples

Author
item LEE, HOYOUNG - Seoul National University
item YANG, CHUN-CHIEH - US Department Of Agriculture (USDA)
item Kim, Moon
item LIM, JONGGUK - National Academy Of Agricultural Science
item CHO, BYOUNG-KWAN - Chungnam National University
item Lefcourt, Alan
item Chao, Kuanglin - Kevin Chao
item EVERARD, COLM - University College Dublin

Submitted to: Journal of Biosystems Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/29/2014
Publication Date: 6/2/2014
Citation: Lee, H., Yang, C., Kim, M.S., Lim, J., Cho, B., Lefcourt, A.M., Chao, K., Everard, C. 2014. A simple multispectral imaging algorithm for detection of defects on red delicious apples. Journal of Biosystems Engineering. 39(2):142-149.

Interpretive Summary: ARS has developed nondestructive spectral imaging technologies for safety and quality evaluations of fruits and vegetables. In this collaborative study, a simple multispectral imaging algorithm for detection of defects on Red Delicious apples was developed using hyperspectral line-scan images. The images were acquired from samples on a moving conveyor at a rate of approximately 4 apples per second through the imaging field of view. Results demonstrated that over 95% of the defects in apples were correctly detected using the spectral algorithm. The detection algorithm can potentially be used on commercial apple processing lines. The method and results presented in this study are beneficial to apple producers and the fruit processing industries.

Technical Abstract: Purpose: A multispectral algorithm for detection and differentiation of defect and normal Red Delicious apples was developed from analysis of a series of hyperspectral line-scan images. Methods: A fast line-scan hyperspectral imaging system mounted on a conventional apple sorting machine was used to capture hyperspectral images of apples moving approximately 4 apples per second conveyor speed. The detection algorithm included an apple segregation method and a threshold function, was developed using three wavebands at 676 nm, 714 nm and 779 nm for execution based on line-by-line image analysis, simulating online real-time line-scan imaging inspection on fruit processing lines. Results: The rapid multispectral algorithm detected over 95% of defect apples and 91% of normal apples under investigation. Conclusions: Use of morphological filtering may enhance the detection rate and reduce false positives. The multispectral defect detection algorithm can potentially be used on commercial apple processing lines.