Location: Quality & Safety Assessment ResearchTitle: Improving performance of real-time multispectral imaging system Author
Submitted to: ASABE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: 6/29/2008
Publication Date: 10/1/2008
Citation: Park, B., Yoon, S.C., Kise, M., Lawrence, K.C., Windham, W.R. 2008. Improving performance of real-time multispectral imaging system. ASABE Annual International Meeting. ASABE Paper No. 085024. Interpretive Summary: ARS scientists have developed real-time multispectral imaging system and tested the performance at the pilot-scale poultry processing line for fecal and ingesta contaminant detection. The imaging system including three-band common aperture camera was able to detect contaminants with high detection accuracy. However, due to time constraints for real-time processing and limited available image processing algorithms, moderate false positive errors were observed. In order to implement the imaging system at commercial poultry processing plant, the false positives must be removed. Therefore, several different approaches in terms of image processing algorithms have been conducted for improving system performance. In this research, noise analyses, optimizing system hardware, gain settings, binning to maximize detection accuracy and minimize false positives were discussed. In addition, the optimum parameter values and the effects of binning relevant to the false positives also were demonstrated. The outcomes from this study could be used for analyzing overall performance of real-time multispectral imaging system for contaminant detection of agricultural commodities.
Technical Abstract: A real-time multispectral imaging system can be a science-based tool for fecal and ingesta contaminant detection during poultry processing. For the implementation of this imaging system at commercial poultry processing plant, false positive errors must be removed. For doing this, we tested and implemented additional image processing algorithms including binning, cuticle removal filter, median filter, and morphological analysis in real-time mode to maximize detection accuracy and minimize false positives (FPs). The median filtering and binning process were able to reduce FPs up to 98.7% and 95.2%, respectively by eliminating most salt and pepper noise from the raw images. The detection accuracy varied with parameter values of image processing algorithms including binning, threshold, median filter, and morphological filter. Overall, detection accuracy for moving birds varied from 84.3% to 97.8%. In this case, the FPs errors were 1.9% and 41.8%, respectively. Although neither the overall detection accuracy nor FPs errors were affected by camera gains, the results of detection accuracy were changed from 87.4% to 95.1%. In this case, the FPs errors were 1.8% and 15.9%, respectively. Thus, the multispectral imaging system was able to detect contaminants with 91.6% accuracy and 3.3% FPs errors by selecting optimum image processing methods at the commercial processing line speed.