Development of Foreign Object Detection Technique at Stage of Receiving Sugar Beet using AI

Abstract

At sugar factories, evaluation of sugar beets is conducted visually by skillful inspectors at receipt stage. Due to labor shortage problem of recent years, automated evaluation technologies for sugar beets have been required. To solve this problem, we developed automated visual evaluation techniques for sugar beets by image processing. In this paper, we report a development of evaluation techniques to detect stone and corrupted sugar beets by using Mask R-CNN.