Classification of good and damaged rice using convolutional neural network

Rice production is massive in Indonesia, therefore maintaining the quality of product necessary. Detection and classification objects have become a very important part image processing. We performed object detection namely rice. After found, it can be classified into two categories, good damaged conducted new study on rice which was carried out per group not grain to obtain or classify where we had several steps, segmentation process using HSV (hue, saturation, value) color space. used because its excellence representing brightness image. considered evaluating tendency darker paler compared To accomodate environment lighting ambiguity perform acquisition controlled environment, so that all images same light intensity. Here use only channel V feature extraction gray-level co-occurrence matrix (GLCM) finally convolutional neural network (CNN) for classification. From test experiments done, produced 83% prediction accuracy. Considering how similar spoiled rice, results are quite impressive.