基于标准粒子群的起重机摄影测量方法

Standard particle swarm optimization based photogrammetry of cranes

大型起重机械的安全性评估与关键尺寸或关键点是高度关联的,文中针对如何高效稳定地获取其三维坐标这一问题,提出一种基于标准粒子群的起重机摄影测量方法。传统相机像主距标定方法主要采用定制的标定板或采用穷举法在一定范围内搜索,但两者的计算精度和计算效率已无法满足目前工程应用中的需求。现提出一种面向起重机的双目相机像主距的标定方法,该方法利用双目测距的数学原理结合粒子群优化算法,将粒子群优化算法(PSO)引入到相机标定中,通过迭代找到最优解,粒子通过自己的经验和同伴中最好的经验来决定下一步的行动。将标准粒子群搜索算法融合到摄影测量方法中,获得精准的有效焦距后,完成起重机关键尺寸的测量。以门式起重机为实验对象进行实验,实验结果表明,相比传统的穷举法,所提方法在速度上具有明显优势,且测量精度能够满足工程需要。

Since the safety evaluation of large crane is closely related to the key dimensions or key points,in this paper,a crane photogrammetry based on standard particle swarm optimization is proposed to figure out how to obtain its three-dimensional coordinates efficiently and stably.In the traditional calibration methods of camera image principal distance,the customized calibration plate or the method of exhaustion are mainly used to search in a certain range,but their calculation accuracy and efficiency are too low to meet the needs of current engineering applications.A method for calibrating the principal distance of binocular camera of cranes is proposed,which combines the mathematical principle of binocular ranging with particle swarm optimization algorithm,and introduces particle swarm optimization algorithm into camera calibration to find the optimal solution through iteration,so that particles can decide the next action through their own experience and the best experience of their peers.After the standard particle swarm optimization(PSO)algorithm is integrated into the photogrammetry method to obtain the accurate effective focal length,key dimensions of the crane are measured.Experiments were carried out with a gantry crane as the object.The results show that compared with the traditional method of exhaustion,the proposed method has obvious advantages in speed and its measurement accuracy can meet the engineering needs.