Skip to main content
ARS Home » Research » Publications at this Location » Publication #171979

Title: REAL-TIME IMAGE PROCESSING FOR RAPID CONTAMINANT DETECTION ON BROILER CARCASSES

Author
item Park, Bosoon
item LAWRENCE, KURT
item WINDHAM, WILLIAM
item SNEAD, MATTHEW

Submitted to: International Society for Optical Engineering
Publication Type: Proceedings
Publication Acceptance Date: 10/22/2004
Publication Date: 11/26/2004
Citation: Park, B., Lawrence, K.C., Windham, W.R., Snead, M.P. 2004. Real-time image processing for rapid contaminant detection on broiler carcasses. International Society for Optical Engineering. 5587:101-111.

Interpretive Summary: Food safety has become an important concern for public health, because reduction in the potential health risks to consumers from human pathogens in food is important food safety issue. While a number of factors can influence bacterial contamination of chicken carcasses, one of the leading causes is fecal contamination at the poultry slaughter houses. Improper plant handling and equipment that is not good working condition can contribute to the problem. Multispectral imaging technique demonstrated potential tools for on-line poultry safety inspection, particularly fecal contamination. Real-time dual-wavelength image processing method was developed for rapid detection of fecal contaminants on broiler carcasses. This new inspection method can improve FSIS poultry safety inspection program, especially HACCP, incorporating scientific testing and systematic prevention of fecal contamination.

Technical Abstract: Recently, the imaging research group at Russell Research Center has developed a real-time multispectral imaging system for fecal and ingesta contaminant detection on broiler carcasses. The prototype system includes a common aperture camera with three optical trim filters (515.4, 566.4 and 631-nm wavelength), which were selected by visible/NIR spectroscopy and validated by a hyperspectral imaging system. This paper demonstrates calibration of multispectral imaging hardware and real-time multispectral image processing software. The software design, especially the Unified Modeling Language (UML) was used for software development. The UML models including class, object, activity, sequence, and collaboration diagram were discussed. Both hardware and software for a real-time fecal and ingesta contaminant detection were tested at the pilot-scale poultry processing line. The multispectral imaging technique can be effectively used for detecting feces and ingesta on the surface of broiler carcasses with a processing speed of 140 birds per minute. The detection accuracy for fecal and ingesta contaminants was approximately 96%. However, the system contains many false positives including scabs, feathers, and boundaries.