Drainage pipeline video data restoration method based on computer vision

Abstract

The invention relates to a pipeline video data restoration method based on computer vision. The method comprises the following steps: carrying out gray stretching on a pipeline image and video collected by a pipeline robot; processing noise interference by using smooth filtering; extracting an iron cable in the center of the video image as a template to realize positioning; performing target identification on the video data center by using an SIFT corner detection algorithm; detecting the cables on the left and right sides of the target by using Hough transform; gray coverage being carried out on the iron cable in the center of the video image and the cables on the two sides; and repairing the data by using an FMM image repairing algorithm. The method can effectively restore the video data collected in the pipeline, reduce the processing difficulty of later pipeline data, improve the processing speed and recognition efficiency of later disease images, and is suitable for the field of urban pipeline and industrial pipeline maintenance.

Description

Technical field [0001] The present invention relates to the field of image and video processing, and more particularly to image repair techniques based on computer vision, including target detection, line detection. Background technique [0002] The detection and repair of pipeline defects is an important part of urban construction and has become a hot research problem of current computer vision. But it is very difficult to get high quality pipeline video data. At present, the pipeline detection mainly relies on the pipeline robot to obtain the internal data of the pipeline, but the mainstream pipeline robot on the market is a driving force of the rope as a power forward, so the rope will inevitably appear in the pipeline video data. The rope will seriously interfere with the recognition of the marginal information defects. In response to the problems present in the above-described pipe video data, the present invention focuses on computer visual video data repair algorithm, usin...

AI Technical Summary

Problems solved by technology

[0003] At present, there are some problems in the field of pipeline defect recognition, such as the interference of the robot's own power device in the pipeline video data, which leads to the fact that in the actual complex pipeline environment, the characteristics of the pipeline defect are easily affected by its changes, and it is impossible to deal with the next pipeline detection work efficiently. , which increases its detection cost

Images