||For the "Le Lectier" (pear), appearance quality is important to determine its grade, and grading based on shipping standards is necessary. However, the workload of grading judgement and the subjective discrimination are problems, and automated and objective discrimination using machines and systems are required to solve the problems. Therefore, in order to develop a pear grading system, a method to detect the appearance quality of pears is investigated. The appearance quality of pears is classified into two major categories: appearance defacement and physiological disorders, and deep learning methods have been used to detect and classify appearance defacement. In this study, we propose a detection method using light image processing based on color features for physiological disorders that have been difficult to annotate due to limited data. The physiological disorders of pears show differences in hue (H), saturation (S) and brightness values (V) in HSV color space compared to normal skin and calyx. The HSV values of these areas are defined as color features, and the physiological disorder areas are detected in the captured images of pears by thresholding using the differences in color features.