Toward Automated Quality Classification via Statistical Modeling of Grain Images for Rice Processing Monitoring

• Main Authors: Jinping Liu, Zhaohui Tang, Qing Chen, Pengfei Xu, Wenzhong Liu, Jianyong Zhu
• Format: Article
• Language: English
• Published: Atlantis Press 2016-01-01
• Series: International Journal of Computational Intelligence Systems
• Subjects: Process monitoring; Sequential fragmentation theory; Weibull distribution; Least squares-support vector machine (LS-SVM)
• Online Access: https://www.atlantis-press.com/article/25868684/view

Computer vision-based rice quality inspection has recently attracted increasing interest in both academic and industrial communities because it is a low-cost tool for fast, non-contact, nondestructive, accurate and objective process monitoring. However, current computer-vision system is far from effective in intelligent perception of complex grainy images, comprised of a large number of local homogeneous particles or fragmentations without obvious foreground and background. We introduce a well known statistical modeling theory of size distribution in comminution processes, sequential fragmentation theory, for the visual analysis of the spatial structure of the complex grainy images. A kind of omnidirectional multi-scale Gaussian derivative filter-based image statistical modeling method is presented to attain omnidirectional structural features of grain images under different observation scales. A modified LS-SVM classifier is subsequently established to automatically identify the processing rice quality. Extensive confirmative and comparative tests indicate the effectiveness and outperformance of the proposed method.