Automatic Inspection Method of Steady Arm Slope Based on Computer Vision

In this paper, a new strategy to inspect the slope of steady arm based on computer vision is presented, in light of the present monitoring ways are not more acceptable for their slowness, imprecision, laborious and potentially hazardous. First, the hardware structure of the image acquisition module of the inspection system is introduced; then, the formulas on calculating the steady arm slope are derivated and the technologies used in image processing are analyzed. Graying and median filter were used in image preprocessing to eliminate noise and enhance image quality; In view of the features of histogram, Otsu threshold segmentation was chose to isolate the registration device from the background; Then, extracting the skeleton of the isolated registration device to improve inspecting precision and finally, using Hough transformation to detect lines and calculate the slope of the steady arm. The experimental result shows that this algorithm is validate.