基于粒子群算法的摄像机自标定  

CAMERA SELF-CALIBRATION BASED ON PARTICLE SWARM OPTIMISATION

作　　者：黄伟光[1] 董安国[1] 

机构地区：[1]长安大学理学院,陕西西安710064

出　　处：《计算机应用与软件》2015年第5期216-219,233,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(11171043;11201038);中央高校基本科研业务费专项资金项目(CHD2012TD015)

摘　　要：基于图像信息,实现对物体的三维重构在交通、地质等领域具有重要的应用价值,对此首先要建立图像坐标和大地坐标的对应关系,而这种关系涉及到摄像机的内部及外部参数,这就需要对摄像机进行标定,确定其参数。利用几何关系给出坐标系间的关系模型,以标定板上关键点间的实际距离和理论距离的相对误差绝对值为目标,将参数确定问题转化为非线性优化问题,进而利用PSO算法对优化模型进行求解,实现对摄像机的自标定。通过实际图像的采集并进行数值计算,结果表明模型正确,与其他算法相比,计算精度得到显著提高。To realise three-dimensional reconstruction of object based on image information has important applied value in the fields of transportation, geology and other sectors. To achieve the purpose, we first establish a corresponding relation between the image coordinates and the geodetic coordinates. Since this relationship is related to internal and external parameters of the camera, we have to do camera cali- bration to determine its parameters. The relational model between the coordinate systems is given by geometrical relationship. Taking the ab- solute value of relative error between the actual distance and the theoretical distance of the key points on calibration board as the target, we convert the parameters determination problem into a nonlinear optimisation problem, and then solve the optimisation model with PSO algorithm and achieve the camera self-calibration. Through the acquisition of actual image followed by numerical calculation, results show that the mod- el is correct, and the accuracy has been significantly improved compared with other algorithms.

关 键 词：摄像机自标定 图像 三维重构 粒子群算法 坐标系 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]