Development of laser sensor sorting system of green soybeans

摘要

The grade-sorting of green soybeans, such as for identifying beans with a defected size, has so far depended on visual inspection and consequently the operation is very time-consuming. It is, therefore, desirable to shorten the operational time requirement through automation. The discernment part which used the laser sensor was built for grade sorting using form information in this research. This thesis reports on the following main results: (1) the development of a laser sensor grade-sorting system, (2) the introduction of a genetic programming of the self-dissembling type, and (3) the establishment of multistage sorting system.