2004 Annual Report
4.What were the most significant accomplishments this past year?
This report serves to document research conducted under a Specific Cooperative Agreement between ARS and the University of Kentucky, Lexington, KY. Additional details of research can be found in the report for the parent project 1265-42000-008-00D "Development of Technology for Automated On-Line Inspection of Animal Carcasses and Plant Produce." A multispectral imaging system and image processing algorithms for identification of wholesome, septicemia/toxemia (septox), and inflammatory process (IP) was developed. Common-aperture multispectral images at three key wavelengths (460, 540, and 700 nm) were collected for 174 wholesome, 75 inflammatory process, and 170 septicemic chickens. The decision tree classification model correctly classified 89.6% of wholesome, 92.3% of inflammatory process, and 94.4% of septicemic chickens. An on-line multispectral classification method to identify systemically diseased chickens was tested on the ISL pilot-scale processing line. Over a period of 6 months, 660 chicken images were collected using a multispectral imaging system with filters at 488 nm, 540 nm, 580 nm, and 610 nm. Classification accuracies of 96% and 97% were achieved for wholesome and systemically diseased chickens, respectively, using image classification features. This simple classification method shows potential for implementation in online in-plant testing in 2005.
|