engleski

An image processing approach for color-based sorting and counting of selected cereals

Physical characteristics of cereals, such as grain size, bulk density, length, shape or color, are of a great importance in terms of product control and quality. Grain counting and sorting is considered as an essential step in product quality assessment, classification and cultivation management in general. However, that procedure is usually carried out manually which can be time-consuming and inconvenient, or by instruments which are rather expensive and bulky. An alternative and nondestructive method for cereal quality evaluation is based on image processing approach, widely used in food industry and agriculture. In this study, an image processing method has been developed and modified for cereal grain counting and sorting based on color segmentation algorithm which allows division of an image into specific regions. The total number of objects in the image can be detected or objects can be classified and counted based on different colors, in this case, using RGB to CIELab color space conversion. Developed method has been tested and evaluted on maize, wheat, buckwheat and rice by varying the number and type of cereal grains for each testing sample. The obtained results showed good efficiency and accuracy of developed image processing method which could be useful counting and sorting tool with a wide range of applications.

cereals ; image processing ; color, sorting ; counting

nije evidentirano