基于亚像素精度分割的晶粒图像识别方法

机器视觉在工业产品质量控制中的应用越来越广,针对LED晶粒的质量检测问题,提出了一种基于亚像素精度阈值分割的方法进行LED晶粒图像的识别。首先通过图像增强及预处理,然后通过基于亚像素精度的阈值分割方法进行晶粒图像的预处理,再通过K值聚类算法进行感兴趣的区域提取,最后通过NCC归一化的方法进行晶粒位置的识别及定位。实验结果表明,提出的亚像素精度阈值分割和NCC归一化定位算法相结合的方法应用在LED分拣机中,能准确地识别出LED晶粒图像。

LED wafer segmented method of image recognition based on sub-pixel precision threshold

Machine vision was used in industrial quality conformity of products more widely.The quality of grain for LED detection,a threshold based on sub-pixel accuracy of the method of segmentation LED grain image recognition was advanced.First through image enhancement and preprocessing,and then based on sub-pixel accuracy by threshold the grain image preprocessing methods,and through the K value of the clustering algorithm extracting the region of interest,and finally through the NCC normalization methods Identification of the location and position of the grain.Experimental results show that the proposed sub-pixel precision threshold segmentation and NCC normalized method of combining localization algorithm used in the LED sorting machine,can accurately identify the LED die image.