基于分块特征匹配的纸箱缺陷检测方法

Defect detection method of carton based on block feature matching

针对物流运输中产生的纸箱缺陷问题,提出一种基于分块特征匹配的纸箱缺陷检测方法。该方法通过将待测纸箱图像与已知图案特征匹配,实现纸箱图案与非图案区域分离;通过Canny边缘检测算法结合图像的矩来计算非图案区域的边缘轮廓面积,判断是否存在缺陷;通过改进的ORB算法结合网格运动统计方法,将图案区域与已知图案的特征点进行匹配并计算匹配率,判定是否存在缺陷。实验表明:该方法检测纸箱图案与非图案区域缺陷的准确率分别为91.5%与93.5%,具有较高的识别率。

A carton defect detection method based on block feature matching is proposed,aiming at the problem of carton defects in transportation.In this method,the carton image to be tested is matched with the known pattern feature to accomplish the separation of the patterned and nonpatterned area of the carton image.The canny edge detection algorithm is combined with the image moment to calculate the edge contour area to identify the defect in the nonpatterned area.The pattern area to be tested is matched with the feature point of known patterns with the improved ORB algorithm combined with the grid motion statistics method,and the matching rate is calculated to identify the defects.The experimental results show that the accuracy of the proposed method is 91.5%and93.5%respectively,which has high identification rate.