Abstract:

Color is an important property that influences the consumers’ purchase decision of a food or agricultural product
and its market value. A non-destructive and rapid detection system for assessing chilled pork quality based on computer
vision system was developed. The working principle, process and hardware composition of the system as well as the
software system were introduced. The hardware system included a pH meter and an image collector, and the acquired images
were processed by binarization, Gauss denoising, edge detection and Hough-transform positioning for the extraction of the
characteristics color parameters by projection algorithm. In addition, pH was used to aid in identifying chilled pork quality.
The experimental results showed that the system could be used for distinguishing chilled pork with high accuracy and
reliability as well as good repeatability.

Key words: chilled pork, distinguishing, pH, colorimetric plate