针对生产过程中产生的一些表面瑕疵会对工业产品质量造成严重威胁的问题,提出一种基于局部二值模式LBP和相位变换相结合的瑕疵检测方法.在分析2种算法优缺点的基础上,针对不同特点的瑕疵采取优势互补策略,提高对各类瑕疵判别、提取和定位的准确性.实验结果表明,该方法检测效果优于采用单独一种方法的结果,对不同种类的瑕疵具有更好的适应性.
Surface defects are inevitably produced during the process of producing industrial products, and some flaws cause serious threats to the quality of products. A defect detection method based on the combination of the local binary pattern (LBP) and the phase-only transform was presented to solve this problem. The advantages and disadvantages of the two algorithms were analyzed, a complementary strategy according to different characteristics of defects was adopted, and the accuracy of the defect identification, extraction and localization was improved. The experimental results show the detection result of this method is better than that of the LBP or the phase-only transform method, and it is adaptable to different kinds of flaws.