Use of Computer Vision during the process of Quality Control in the classification of grain

Abstract

In this work the date collected is a result of the development of a system of computer vision taking place during the quality control stage of the process of classifying grain sizes. This is important as the machines may need to be adjusted if samples of produce are found to have differing percentages of grain quality during the process. This study focussed on the machine process involved in the production of rice for small and medium-sized companies, referred to in this study as S&MC. This production process is made up of different stages; the first is the acquisition of images of the grains of rice which are then condensed to obtain better results in the segmentation process. The segmentation process uses images which seek the outline of the grains only. This study took these images and later developed them further and algorithms were programmed in order to segment the whole rice grains and those three quarter size, using techniques of characterization and grain classification based on the morphologic properties of these grains. Finally, it was possible to segment and to classify 90.92% of the images analyzed in the correct form.