17 March 2017 Classification of rice grain varieties arranged in scattered and heap fashion using image processing
Author Affiliations +
Proceedings Volume 10341, Ninth International Conference on Machine Vision (ICMV 2016); 1034126 (2017) https://doi.org/10.1117/12.2268802
Event: Ninth International Conference on Machine Vision, 2016, Nice, France
Inspection and classification of food grains is a manual process in many of the food grain processing industries. Automation of such a process is going to be beneficial for industries facing shortage of skilled workforce. Machine Vision techniques are some of the popular approaches for developing such automations. Most of the existing works on the topic deal with identification of the rice variety by analyzing images of well separated and isolated rice grains from which a lot of geometrical features can be extracted. This paper proposes techniques to estimate geometrical parameters from the images of scattered as well as heaped rice grains where the grain boundaries are not clearly identifiable. A methodology based on convexity is proposed to separate touching rice grains in the scattered rice grain images and get their geometrical parameters. And in case of heaped arrangement a Pixel-Distance Contribution Function is defined and is used to get points inside rice grains and then to find the boundary points of rice grains. These points are fit with the equation of an ellipse to estimate their lengths and breadths. The proposed techniques are applied on images of scattered and heaped rice grains of different varieties. It is shown that each variety gives a unique set of results.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Sudhanva Bhat, Sudhanva Bhat, Sreedath Panat, Sreedath Panat, Arunachalam N, Arunachalam N, } "Classification of rice grain varieties arranged in scattered and heap fashion using image processing", Proc. SPIE 10341, Ninth International Conference on Machine Vision (ICMV 2016), 1034126 (17 March 2017); doi: 10.1117/12.2268802; https://doi.org/10.1117/12.2268802

Back to Top