An Economical Design of Automatic Rice Grading using Image Processing

Abstract

Product quality inspection is a crucial step in the production line of rice industries. To maintain the quality and enhance the inspection methodology, the challenge is to scrutinize rice grain individually over entire batch for escalating the yield. The conventional method used in rice industries is to examine the quality of rice grains manually. The decisions made by human quality control inspectors may be affected by external influences like fatigue, exhaustion or stress which causes non uniformity in evaluation procedure and generate the high probability of errors. The major drawback of the manual inspection comprises high labor content and expenditures. This research paper provides a cost effective design solution to overcome the described limitations by developing a system which helps in sorting rice and eliminate the manual examination. For attaining the automatic grading of rice, the image processing technique is applied which help to sort defected rice grain from the entire batch of rice grains. The system has been developed on MATLAB which helps in the inspection process and its graphical user interface provides information of rice grains in three different quality based categories which sorted on the basis of size and colors. To identify the defected grains, the multilevel threshold method of image processing has been used. The proposed design also helps to determine the quantity of defected rice grains by evaluating the quantity of discolored grains and to identify the size of the rice grain, the geometrical features extracted for each individual rice grain are used to estimate the length