大米粒形边缘检测算法研究

Study in rice figure edge detection algorithms

目前,大米品种主要由人工来检测,由于人为因素的影响,造成检测结果主观性强,一致性差。为了解决这一问题,可以利用计算机对大米品种外形进行检测。基于此,讨论了采用几种不同的边缘检测算子来实现对大米外形的检测并比较其结果。实验表明,该方法用于大米粒形的识别时对大米的判定正确率较高。

Now the rice variety had been identified mainly by human inspection. Due to artificial factor, the result of inspection was subjective and non-identical. In order to solve this problem, the rice variety could be identified by computer vision. So some different edge detection operators were presented to complete detection for the rice figure and compare their effects. The experiment result indicates that the correctness is comparatively high while identifying the rice figure by the method.