The Recognition and Localization of Insulators Based on SIFT and RANSAC

The recognition and localization of insulators is a crucial process of insulators' failure detection, especially under the complex background. With unknown insulators in test image, a method of the recognition and localization of insulators based on SIFT and RANSAC is proposed as following: SIFT (Scale Invariant Feature Transform) is used to extract a set of feature points of each insulator template image and test image. Then use RANSAC (RANdom SAmple Consensus) to remove the outliers. And each template's matched inliers are accurate points which are used to compute affine transform matrix H and localize the template images in test image. Thus we can detect all the template images in test image correctly. Experimental results show that the proposed approach can obtain high precision in the recognition and localization of insulators under the complex background.