The growth of unacceptable molds on Kashkaval cheese is a quality problem that can be caused by various factors relating to storage conditions and packaging. Computer vision methods are preferred for solving many food quality problems due to their benefits, such as fast, non-destructive evaluation and low- cost equipment. Thus, the current research focuses on the opportunity for computer-based identification of areas containing mold on the surface of Kashkaval cheese using images processing in HSI color space. Since the predominant mold species on Kashkaval cheese is white, its detection poses a significant challenge for conventional image processing methods based on thresholding or segmentation. Regarding this, the current research investigates the effectiveness of segmentation in the HSI color space together with the prioritization of color components. A set of images manually processed by experts is used to assess the automatic localization of mold. Based on arithmetic and logical operations, segmented images of Kashkaval cheese are compared with those of the corresponding manually prepared set, and the results indicate that the prioritization in HSI color space is relevant to the task of automatic mold identification.
Bosakova-Ardenska et al. (Mon,) studied this question.