Submitted to: American Society of Agricultural Engineers Meetings Papers
Publication Type: Proceedings
Publication Acceptance Date: July 18, 2005
Publication Date: August 2, 2005
Citation: Yoon, S.C., Park, B., Lawrence, K.C., Windham, W.R. 2005. Statistical modeling of multispectral images for improved contamination detection on poultry carcasses. American Society of Agricultural Engineers Meetings Papers. Paper No. 053072. Interpretive Summary: Developing a detection technology to find the presence or absence of fecal contamination on the surface of poultry carcasses is critical to food safety. A statistical detection technique was developed to discriminate between a contaminated bird and an uncontaminated bird. Multispectral image data of fecal material and uncontaminated normal carcasses was statistically modeled with two distribution estimators. The study found that the statistical population of fecal material was not normally distributed and that of uncontaminated carcasses was normally distributed. A projection was used to reduce the variation of multispectral data and to model the statistical characteristics. A linear mixture of statistical distribution estimates was designed to develop a fecal material detection algorithm. This algorithm can be incorporated into a real-time multispectral imaging system to detect visible fecal contaminants on boiler carcasses.
Technical Abstract: Developing a detection algorithm to decide the presence or absence of fecal contamination on the surface of poultry carcasses is critical to food safety. The single threshold strategy for a band ratio algorithm has been known to be limited to pixel-basis detection. In an attempt to develop a statistical rule for carcass-basis detection from multispectral images, probability density functions of both contaminated and uncontaminated materials were estimated by parametric and nonparametric methods. We found that uncontaminated poultry carcasses could be modeled by a Gaussian distribution, whereas contaminated materials were non-Gaussian. A kernel density estimator was used to analyze the non-Gaussian characteristic of the fecal material on a transformed projection axis. A linear mixture of the density functions was introduced to model the observations made on the projection axis. A new detection algorithm was designed using the mixture model and tested for 102 birds (56 dirty and 56 clean birds). A preliminary test on the sample birds revealed that the algorithm needed at least 12 contaminated pixels to reach the perfect detection results. The test also showed a false positive rate of less than 5%.