Research on feature points extraction and matching technology of pipe-wall images

Abstract

Pipeline defect recognition is necessary to achieve the online inspection of pipeline. In this paper, the method of pipeline visual inspection is proposed. According to the principle of pipe imaging, Harris corner extraction operator is used to present the NCC algorithm of corner detection and automatic point matching. The feature point identification and matching algorithm of pipeline defect is studied, and the corresponding analytical result indicate the extraction and matching algorithm of pipeline visual inspection is effective.