首页|Online double-sided identification and eliminating system of unclosed-glumes rice seed based on machine vision
Online double-sided identification and eliminating system of unclosed-glumes rice seed based on machine vision
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
An online double-sided identification and eliminating system based on machine vision was developed to identify the unclosed-glumes rice seed by analyzing its double-sided image simultaneously. The vibrating plate and linear vibration conveyor achieved the transfer of rice seed from disordering to spacing, and the identified unclosedglumes rice seed was eliminated by the blowing jet. Hough linear detection and extracted feature were used to identify the unclosed-glumes rice seeds. The algorithm achieved the accuracy of 88.1% for normal seeds and 87.7% for unclosed-glumes seeds when using double-sided image acquired online. Multi-thread processing was used for double-sided images to shorten the code execution time. The online double-sided identification and eliminating system achieved the average accuracy of 83.7% for normal seeds and 83.3% for unclosed-glumes seeds. Results show that the system has a good adaptation of different rice seed varieties and achieved desired accuracy for online identification and eliminating of unclosed-glumes rice seed.
NSTL
Elsevier
