Submitted to: Near Infrared Spectroscopy News
Publication Type: Trade journal
Publication Acceptance Date: 9/4/2001
Publication Date: 10/1/2001
Citation: Lawrence, K.C., Windham, W.R., Park, B., Buhr, R.J. 2001. Hyperspectral imaging for poultry contaminant detection. Near Infrared Spectroscopy News. Interpretive Summary: In a poultry processing plant, where carcasses travels on processing- line shackles at a speed of 140 birds per minute, a potential source of pathogen contamination on a carcass is poultry feces and ingesta. We are developing a research imaging system to check for fecal or ingesta (undigested food) contamination on the carcasses. Because the shackle line speed will limit our final image processing time, and the complexity of the problem, specific images must be collected that optimize the contrast between the contaminant and the rest of the carcasses. This paper reports the initial results that identify key wavelengths in the visible light range for detecting contaminants. These key-wavelength were then tested on images of whole carcasses contaminated with small spots of feces and ingesta. A detection algorithm was also developed that enhances the separation between the carcass and the contaminants. Results show that 100% of the contaminant spots were detected. More work is needed to see if the chicken's diet and other processing variables, such as scald-water temperature, affect the results. These wavelengths and detection algorithms will later be applied in an imaging camera with special filters so that a final system can operate at processing line speeds.
Technical Abstract: Contamination of meat and poultry with bacterial food-borne pathogens can potentially occur as a result of exposure of the carcass to ingesta and/or fecal material during or after slaughter. A method and system for detecting fecal and ingesta contaminates was demonstrated. Visible / Near Infrared reflectance data were first used on fecal and uncontaminated skin samples to determine four key wavelengths for evaluation with a hyperspectral imaging system. Single key wavelengths were found to be inadequate for detecting contaminates. However, a ratio of an image at 565 nm divided by a 517-nm image was best for identification of fecal and ingesta contaminates. The ratio image was then further processed by masking the background and enhancing the contrast with a nonlinear histogram stretch. The results indicated that, for the limited sample population, 100% of contaminates were detected. Further research needs to be conducted to evaluate the effect of diet on fecal contaminates and the effect scalding procedures on the carcass skin and to transfer the technology to a real-time imaging system.