Using Digital Image to Classify Phenotype of the Rice Grain Quality under Agricultural Standards Act

摘要

Digital image analysis (DIA) was widely used to describe plant organ shape. However, the various types of shape descriptors that can be generated through the DIA were certainly applicable to identify different size of rice grain. In that, it helped the nice quality inspection more convenient and reduced fatigue from work. Concurrently, the results must be close to human work. As to the results from the experiment, it resealed grain size distribution according to the Agricultural Product Standards ACT, was averagely 5 percent error, and 0.031 percent error when increased against human performance and tools NIR-SC-5000-1/2In views of processing speed, the results of the hands-on experiment took an average of 5.20 minutes per 10 seeds. For a program developed at an average speed of 19 seconds per 10 seeds, it was found that digital imaging was used to measure grain size. Phenotypic accuracy was no different from the expert. It also more resolution than that of the tools used now a day revealed.