Application of Image Segmentation Technique to Quantify Cleanliness Value of Peeled Water Chestnut

Abstract

In this research paper, in order to gain assessment on the quality of peeled water chestnut, the optimization of the water chestnut peeling machine is primarily conducted via the peeling mechanism, which is the rotating cutter assembled on the top of the machine. The optimal cutting parameters' conditions, including cutting angle, cutting height, cutting velocity and peeling du-ration are determined. The quality parameters representative in this experiment are the two characteristics: Cleanliness Value (CV) and Remaining Weight. As regards the method of obtaining CV, in the first stage, raw image of peeled water chestnut is captured by digital camera. After that, the image is transformed to 8-bit Grey Scale by applying image processing technique. In the following step, image segmentation method is utilized in order to extract features of image. Meanwhile, the thresholding algorithm is applied for unpeeled skin and full outer boundary of water chestnut. Finally, the CV can be obtained by the ratio of pixel area of unpeeled skin to area of full outer boundary.