Multi-spectral imaging of food quality in thermal processing

Thermal processing is recognized as a major source of quality degradation of heat-sensitive foods, resulting in almost 95% loss of natural vitamins and bioactive components (Sun, 2006). The challenge of maintaining both quality and safety in thermal processing depends on our understanding of mechanisms of quality degradation. Due to the number of physical and biochemical mechanisms involved, there is a gap in our knowledge of temperature-induced quality degradation. Although the sensory evaluation of quality is appropriate and may be the most accurate method, sensory panels are not always available for routine, large-scale evaluation of food quality. The tremendous resolution of computer vision has opened the way for real-time imaging of food quality degradation. However, the use of computer vision for real-time imaging of food quality has been limited because of the difficulties in interpretation of quality indicators from digital images. The objective of this proposed research is to establish relationships between food quality indicators and morphological, colour and textural attributes of digital images. A high-resoluton computer vision system will be used for real-time imaging of heat-sensitive food. Simultaneous multi-spectral imaging of colour and temperature in the process of heating will add considerable knowledge to our current understanding of temperature-induced quality degradation. This technology is not in common use and its development will create opportunities for research and industrial applications of "smart" engineering tools in the food industry.