Abstract
The sorting rate of green soy bean pods has remained low because the work generally has to be done by hand. Therefore, producers have expressed interest in a grading machine. In the present study, we investigated whole surface image capture technology of green soy bean pods and developed an experimental image capture machine. The bean pods were dropped from a conveyor belt and photographed in mid-air before two mirrors set at a 120 degree angle to one another, allowing the camera to view the entire surface of each bean pod. Moreover, the bean pods always fell at an incline, and a posture stability board made them parallel with the direction of their fall. In trial tests, sorting accuracy for the image capture machine was 79 percent.
