鲜杏外部缺陷在线检测系统

On-line Detection System of Fresh Apricot External Defect

为了有效实施对鲜杏表面缺陷的快速检测,基于LabVIEW软件平台和IMAQ Vision视觉工具包,开发了可同时检测鲜杏表面磨伤、日灼、霉变及碰伤缺陷的在线视觉检测系统。该系统包含图像在线采集和处理识别模块。最后,通过对搭建的系统在水平输送速度为0.063m/s条件下,对190个各类鲜杏样本进行RGB彩色图像采集,然后提取G单分量图像并依次采用中值滤波、图像灰度变换、阈值分割及形态学处理等算法对缺陷鲜杏图像进行处理和分析,得到正常果、磨伤果、日灼果、霉变果、碰伤果检测的准确率分别为88.00%、73.33%、80.00%、8 2.5 0%、8 7.5 0%,整体在线识别率达到8 3.1 5%。

To detect external defects of fresh apricot, on-line inspection systems was developed based on LabVIEW and IMAQ Vision for detection of fresh apricot external defects, such as surface rubbed,sun-burn, mildew and bruised. In this study, the speed of the convey belt of the system was set as 0. 063m/s. Under this condition fresh apricot RGB images were acquired. Then, image of G-component was extracted from the RGB images, and then apricot external defects were identified using the following image processing methods orderly median filtering, gray transform, threshold and morphological processing. The test results showed that detection accuracies of the sound apricots, rubbed apricots, sun-burn apricots, mildew apricots, bruised apricots were 88.00% ,73.33% , 80.00% , 82.50% and 87.50% respectively, integrated recognition accuracy was 83.15 %.