Center of circle positioning method of heat pipe orifice based on random hough transform

author: Yong Li, Haoting Han, Tengyuan Liu, Bin He, Jiabin He, Yuanzhe Liu
status: in the first round of peer review
abstract:
In order to solve the problems of burr, depressions and out of roundness in the actual production process of existing heat pipes, a center of circle positioning and detection method based on random Hough transform is proposed. Firstly, the single pixel edge of the heat pipe orifice image is obtained by image preprocessing, minimum circumscribed circle fitting and morphological edge detection. After that, a random circle is fitted by three points, and a threshold is set to determine whether the boundary pixels are located on the circle or not. Finally, the percentage between the point on the circle and all the boundary pixels are calculated to determine the center of the circle. It effectively solves the problem of interference with the positioning of the center of the circle due to the presence of burrs or out of roundness phenomena near the orifice. Meanwhile, the simulated orifice images are used for testing, and the theoretical fitting error is within the range of 3 pixels, which has a very good robustness. The experimental results show that for 2~6mm heat pipe orifices, the algorithm in the paper, compared with the Tukey-based iterative weighted least squares fitting algorithm, has the deviation of horizontal and vertical coordinates within 0.0257mm, the deviation of radius within 0.0183mm, and the running time within 0.5s, which can meet the requirements of the current production line.