Recognition of wrinkle black soybean by machine vision

Abstract

It is useful to use a machine vision in order to measure the outside shape and the surface texture of an object.This research is a fundamental study for a judgment of outside quality for the black soybean and objective is to develop an algorithm for the judgment by using the machine vision.In this report, a kind of damaged black soybean, that is a wrinkle soybean, was focused on the subject for the judgment.The wrinkle part of the black soybean was extracted by the two dimensional Fast Fourier Transform (FFT) method.After processing the original image by FFT, a two dimensional power spectrum was obtained and was filtered out with two range of frequency from 10 to 20 (cycle/pixel) or from 30 to 40 (cycle/ pixel).The filtered power spectrum was processed by inverse FFT and then the wrinkle part was extracted.There was a remarkable difference between the processed image of the wrinkle and whole soybeans.A percentage of judgment error of the developed algorithm for the wrinkle black soybean was 35 % in 30 samples.

Abstract from the paper’s OpenAlex record, shown for indexing and research. Copyright remains with the authors or the publisher. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.