Figure 16. Schematic diagram of industrial camera layout

3.2 Evaluation indicators for testing results

In this paper, the average distance between the intersection coordinates between the detection lines and the corresponding corner coordinates of the suspender is used to evaluate the error size of the line detection results, and determine whether the line detection results are within the error range. Manually mark A, B, and C as the suspender corner points, the intersection point between line and line is a, the intersection point between line and is b, and the intersection point between line and is c. Aa, Bb, and Cc are error distances between corresponding coordinate points. Finally, the average error distance of Aa, Bb, and Cc is used to measure the detection result. The schematic diagram of evaluation criteria is shown in Figure 17:
Figure 17. Schematic Diagram of Evaluation Criteria

3.2 Analysis of results under different levels of interference

This article lists four datasets with different interferences, and uses the method proposed in this article and the original k-means clustering method to detect four datasets. Each example includes an edge detection graph, a Hough transform graph, the detection results of the two algorithms, and a scatter plot of the straight line voting numbers of the clustering results. To facilitate the display of the detection effect, the detection image is now magnified by 10 times, and the test example is shown in Figure 18.