为了对大尺寸筒状设备进行圆度误差评定，研制了大尺寸筒状设备圆度误差测量系统。其结合了激光准直、图像处理等多方面技术对大尺寸筒状设备的圆度误差评定进行研究。首先，介绍了系统获取大尺寸筒状设备圆度数据的方式。然后，根据圆度数据的获取方式，提出了基于相邻测点位置关系的最小外接圆过滤算法。最后，验证该算法的过滤效率，并与同类算法在解算时间、解算结果两方面进行比较。实验结果表明：最小外接圆过滤算法解算时间较以往同类算法缩短了30%以上。系统CCD测量杆的测量精度达到0.7 mm。大尺寸筒状设备圆度误差测量系统满足工程需要，可对大尺寸筒状设备进行有效圆度误差评定。

In order to evaluate the roundness error of large-scale sleeve, a roundness error measurement system of large-scale sleeve was established and its applied algorithms such as laser collimation, image processing and etc were investigated. First, the method that the system obtained roundness error of large-scale sleeve was introduced. Then, based on the measuring process, a filtering algorithm of minimum circumscribed circle which analyzed the point position forms was presented. Finally, filtering efficiency was analyzed. Computing time and the precision were compared with the similar algorithm. Experimental results indicate that the computing time of the filtering algorithm is shorter by more than 30% and the precision of the CCD measurement rod is 0.7 mm. The roundness error measurement system of large-scale sleeve can satisfy the requirement of the project and evaluate roundness error effectively.