Skip to main content
ARS Home » Research » Publications at this Location » Publication #156393
Title: DEVELOPMENT OF MULTISPECTERAL IMAGING PROCESSING ALGORITHMS FOR IDENTIFICATION OF WHOLESOME, SEPTICEMIA, AND INFLAMMATORY PROCESS CHICKENS

Author
item YANG, C - UNIV OF KENTUCKY
item Chao, Kuanglin - Kevin Chao
item Chen, Yud

Submitted to: Journal of Food Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/16/2004
Publication Date: 7/21/2005
Citation: Yang, C., Chao, K., Chen, Y.R. 2005. Development of multispecteral imaging processing algorithms for identification of wholesome, septicemia, and inflammatory process chickens. Journal of Food Engineering. 69(2):225-234.

Interpretive Summary: The Poultry Products Inspection Act (PPIA) requires Food Safety and Inspection Service (FSIS) inspectors of the United States Department of Agriculture (USDA) to conduct post-mortem inspection for wholesomeness of all chickens intended for sale to U.S. consumers. FSIS has just completed a 3-year transformation of its traditional inspection system to a Hazard-Analysis-and-Critical-Control-Point (HACCP) inspection system. Under HACCP, increasing consumer demand and line speeds will continue to increase the need for and pressure on inspectors. Thus, to address food safety concerns and meet growing consumer demand, there is an urgent need to develop automated inspection systems that can operate on-line in real-time in the slaughter plant environment. The objective of this study was to develop multispectral image algorithms for the detection of several categories of infectious poultry conditions. Multispectral images for 174 wholesome, 170 septicemia, and 75 inflammatory process chickens were collected using a common-aperture camera with interference filters at 460 nm, 540 nm, and 700 nm. Two image processing algorithms were developed for the specific identification of septicemia and inflammatory process chickens. After principal component analysis was applied to the multispectral images, separation of septicemia chickens was found possible by calculating the average intensity of the first principal component images. With a threshold value of 105, the average intensity value for 95% of the septicemia chickens was below that threshold and thus those chickens could be correctly differentiated from wholesome and inflammatory process chickens. For separating inflammatory process chickens from wholesome and septicemia chickens, a region of interest was defined around the lower abdomen in the multispectral chicken images and spectral features for inflammatory process were determined. All inflammatory process chickens were successfully separated by the algorithm's identification of pixels satisfying those spectral feature conditions. A decision tree model was constructed to classify chickens based on input from both algorithms, and was able to correctly classify 89.6% of wholesome, 94.4% of septicemia, and 92.3% of inflammatory process chickens. This information is useful to the Food Safety and Inspection Service (FSIS), and poultry processing plants.

Technical Abstract: A multispectral imaging system and image processing algorithms for food safety inspection of poultry carcasses were demonstrated. Three key wavelengths of 460, 540, and 700 nm, previously identified using a visible/near-infrared spectrophotometer, were implemented in a common-aperture multispectral imaging system, and images were collected for 174 wholesome, 75 inflammatory process, and 170 septicemia chickens. Principal component analysis was used to develop an algorithm for separating septicemia chickens from wholesome and IP chickens based on average intensity of first component images. A threshold value of 105 was able to correctly separate 95.6% of septicemia chickens. For an algorithm to differentiate inflammatory process chickens, a region of interest was defined and spectral features to identify inflammatory process pixels were determined. The algorithm was able to correctly identify 100% of inflammatory process chickens by detecting pixels that satisfied the spectral feature conditions. To combine both methods, a decision tree model was created to classify the three chicken conditions using inputs from the two image processing algorithms. The results showed that 89.6% of wholesome, 92.3% of inflammatory process, and 94.4% of septicemia chickens were correctly classified.