Integrated Circuit Image Feature Extraction Based on K-Means Clustering Combined with Harris

To effectively extract corner point feature information from IC chip images, a technique of extracting corner point feature information from IC defective chip images combining K-means clustering algorithm and Harris algorithm is proposed. Firstly, the traditional Harris algorithm is improved from the perspective of adapting to the processing of IC chip images; secondly, the IC chip images are detected using Harris corner point detection and then combined with K-means based clustering algorithm to achieve the optimal detection effect of the image target; finally, the proposed algorithm is compared with Moravec algorithm, ORB algorithm, SIFT algorithm and Finally, the proposed algorithm is compared with Moravec, ORB, SIFT and traditional Harris corner detection algorithms, and the performance of the algorithm is further explained by objective evaluation using direct statistics and common evaluation indexes such as the running time of the algorithm. The experimental results show that this method can achieve the corner point feature detection of IC chips, so that the obtained feature areas not only contain as many stable corner points as possible, but also do not overlap each other, and the method can effectively extract the real corner points in the image, which has strong applicability.