Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

This research was performed to develop an algorithm and machine vision system for classifying appearance characteristics of rice, namely head rice, cracked rice, chalky rice, broken rice, and milling rate of rice, whether kernels were isolated or grouped. The capability of developed machine vision system for accurate and rapid identification of bulk samples was tested. The developed algorithm in this study was able to correctly classifying rice features. For segmentation, an overall accuracy of 98.9% was achieved. For milling rate, broken rice, cracked rice, and chalky rice, overall accuracy were 99.6%, 99.5%, 97.6%, and 94.7% respectively. It took 2 seconds (include rice supplying, image acquisition, image analysis, and rice discharging) to classify 200 kernels at a time using developed machine vision system. It was possible to calculate the average milling rate of total sample and the milling rate of each single kernel. From this result, it was considered that the developed system was useful for comparing the performance of a rice whitener. It was concluded that the developed system was enough to use for inspection of appearance quality of rice.