Submitted to: Computer Vision Technology for Food Quality Evaluation
Publication Type: Book / Chapter
Publication Acceptance Date: 11/19/2015
Publication Date: 5/9/2016
Citation: Park, B. 2016. Quality evaluation of poultry carcasses. Computer Vision Technology for Food Quality Evaluation. Chapter 9: pp 213-218.
Interpretive Summary: In the United States, poultry has the highest per capita consumption of all meats. Ensuring that it is safe to eat is important to both producers and consumers. Poultry meat uptake per person increases faster than pork and beef with an estimated 9% rise over the next decade compared to gains of approximately 3~4% for beef and pork meat. Thus poultry products keep increasing in popularity with consumers and retail values in the U.S. broiler industry. To maintain the trends of increasing poultry consumption, the poultry industry is always concerned with maintaining quality control, as well as standard quality evaluations which are extremely important to both poultry producers and consumers. The USDA has been mandated to inspect poultry carcasses on-line at processing plants. In doing this, the development of accurate and reliable methods for detection of unwholesome poultry carcasses is essential to improve quality as well as safety inspections. Machine visions with appropriate image processing methods have the potential for automated inspection and evaluation of poultry quality and safety. In this article, machine vision and imagine processing methods with their applications for poultry industry are discussed.
Technical Abstract: The USDA Food Safety Inspection Service (FSIS) has been mandated to organoleptically inspect poultry carcasses online at processing plants. For poultry quality and safety evaluation, the development of accurate and reliable instruments for online detection of unwholesomeness such as septicemia, cadaver, bruised, tumorous, air-sacculitic, and ascites poultry carcasses are essential to improve the task for quality inspection. Machine vision is promising technology for quality control by separating unwholesome carcasses during poultry processing, particularly in tedious and repetitive processing for grading and inspection. According to poultry inspection practice, poultry carcasses are visually inspected by inspectors at the poultry processing plant. Since visual bird-by-bird inspection is labour-intensive and prone to human error, high speed and reliable inspection systems to ensure safe production of poultry has become an important issue, as the public is demanding assurance of better and safer food. In this book chapter, several machine vision methods including color and spectral imaging such as hyperspectral and real-time multispectral imaging are demonstrated for poultry quality evaluation. In addition, recently developed imaging platforms such as real-time hyperspectral imaging, transportable multispectral imaging, handheld multispectral and hyperspectral microscope imaging technology are discussed for poultry industry applications.