基于改进智能霍夫变换的红热圆形工件直径检测  

An improved intelligent Hough transform for detecting the diameter of red-hot circular workpiece

作　　者：张新雨[1,2,3] 李思雨 李婧华 杨延西 ZHANG Xinyu;LI Siyu;LI Jinghua;YANG Yanxi(National&Local Joint Engineering Research Center of Crystal Growth Equipment and System Integration,Xi’an University of Technology,Xi’an 710048,China;Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing,Xi’an 710048,China;Faculty of Automation and Information Engineering,Xi’an University of Technology,Xi’an 710048,China)

机构地区：[1]西安理工大学晶体生长设备及系统集成国家地方联合工程研究中心,陕西西安710048 [2]陕西省复杂系统控制与智能信息处理重点实验室,陕西西安710048 [3]西安理工大学自动化与信息工程学院,陕西西安710048

出　　处：《西安理工大学学报》2022年第1期41-47,共7页Journal of Xi'an University of Technology

基　　金：国家自然科学基金重点资助项目(61533014);国家自然科学基金资助项目(62003261,61702410,62073258);陕西省自然科学基金资助项目(2020JQ-645)。

摘　　要：为提高霍夫变换(Hough transform,HT)检测圆形工件直径的速度,本文提出了一种基于模拟退火(simulated annealing,SA)的改进智能霍夫变换方法。首先将直径检测问题转化为在椭圆参数空间中搜索最优解的问题,并根据霍夫变换拟合椭圆的原理构造了目标函数。然后针对椭圆检测问题设计了混合Metropolis Hastings(MH)抽样算法,对模拟退火算法中的抽样方法进行改进,最后利用改进的模拟退火算法在参数空间中搜索最优椭圆参数,得到圆形工件的直径测量值,从而减少了霍夫变换算法的计算量和内存占用量,提高了霍夫变换的速度,并通过仿真和实际实验验证本文方法的有效性。实验结果表明,与迭代最小二乘(iterative least square,ILS)、随机霍夫变换(random Hough transform,RHT)和随机抽样一致(random sample consensus,RANSAC)相比,本文方法能够快速准确地检测椭圆参数和红热圆形工件直径。To improve the speed of Hough transform(HT) in the diameter detection of circular workpiece, an improved intelligent Hough transform method based on the simulated annealing(SA) is proposed in this paper. Firstly, the problem of the diameter detection is transformed into that of searching the optimal solution in the ellipse parameter space, with the objective function constructed according to the principle of fitting ellipse by HT. Then, a mixed Metropolis Hastings(MH) sampling strategy is designed for the ellipse detection, with the sampling method of simulated annealing algorithm improved. Finally, the improved simulated annealing algorithm is used to search the optimal ellipse parameters in the parameter space to obtain the diameter measurement of the circular workpiece, thus reducing the calculation and memory consumption of HT. The speed of HT is improved significantly, and the effectiveness of the proposed method is verified by simulation and actual experiments. Experimental results show that, compared with the iterative least square(ILS), the random Hough transform(RHT) and the random sample consensus(RANSAC), the proposed method can detect ellipse parameters and red-hot circular workpiece diameter quickly and accurately.

关 键 词：智能霍夫变换 椭圆拟合 模拟退火 直径检测 圆形工件 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置] TP29[自动化与计算机技术—控制科学与工程]