Study on Detection Method of External Defects of Potato Image in Visible Light Environment

Potato as the fourth largest staple food in China, The external defect detection directly affects the industrialization of potato and deep processing. As the currently domestic testing method are mostly based on specific circumstances, specific light, which does not satisfy the testing requirements of actual environment. Therefore, this paper presents a non-destructive method for the study of green, germination and lesion of potatoes in the visible environment, which has a great significance for the deep processing and commercialization of potato. In this paper, firstly, we studied the segmentation method of potato image in visible light environment and proposed a new method to split the potato target area, subsequently, we respectively studied the detection method of defect area. For the green skin region, a new detection method based on RGB, HSV and LAB multi-color model was proposed. A method based on Laplace operator's gray variance is proposed for the germination and lesion area. The experimental results showed that the proposed method is effective for our study.