4d.Progress report.
This report documents research conducted under a specific cooperative agreement between ARS and the University of Kentucky. Additional details of research can be found in the report for the in-house associated project previously #1265-42000-008-00D entitled "Development of Technology for Automated On-Line Inspection of Animal Carcasses and Plant Produce," and currently #1265-42000-013-00D entitled "Development of New and Improved Systems to Enhance Food Safety Inspection and Sanitation of Food Processing."

A fast line-scan imaging system was developed and used to take images of 113 wholesome and 114 systemically diseased chickens in two sets: 65 wholesome and 74 systemically diseased for the first image set, and 48 wholesome and 42 systemically diseased for the second image set. The chicken carcasses were hung on a pilot-scale shackle moving at a speed of 70 carcasses per minute. The light-emitting-diode (LED) line lights were selected as the appropriate light source since the resulting difference in relative reflectance between wholesome and systemically diseased chickens is more significant and the reflectance intensity for shorter wavelengths is higher than that using other light sources. Among the 103 available wavelengths from 395 nm to 1138 nm, four wavelengths at 413 nm, 472 nm, 515 nm, and 546 nm were selected through spectral analysis as key wavelengths when using the LED line lights. The wavelength of 626 nm was selected as the reference wavelength to obtain the ratio of relative reflectance between each key wavelength and the reference wavelength. A fuzzy logic based algorithm was developed to differentiate images between wholesome and systemically diseased chickens. For each scanned line, four image features from single pixels were used as inputs to run the fuzzy logic algorithm to obtain a discrete decision output, indicating the existence of systemic disease. Two methods were investigated in this study. For the first method, the average decision output of pixels on the chicken surface of all scanned lines in the region of interest was analyzed. The successful result showed that with proper selection of key wavelengths and image features, the fuzzy logic algorithm could quickly and accurately differentiate systemically diseased chickens from wholesome ones. The result showed that when the average pixel decision output was higher than 0.50, the chicken could be condemned as being systemically diseased. For the second method, the average decision output of pixels from chicken surface of each scanned line in the region of interest was analyzed, and the lines in which the average output was higher than 0.50 were counted. The result showed that when the count of lines suspected of systemic disease was higher than 40, the carcass could be condemned as being systemically diseased, indicating that scanning of a whole carcass may not be necessary to make a final decision on that carcass.