一种视频图像规则圆目标的超分辨率位置测量方法

介绍了一种用于视频图像规则圆目标的超分辨率位置测量方法。分析了超分辨率测量的原理 ,并提出和完成了测量方案 ,取得了满意的结果。首先在目标静止时 ,连续采集多帧视频图像 ,通过叠加平均进行降噪 ;其次通过罗伯特微分算子进行边界探测 ,求出目标的大致位置 ;最后通过边界重心拟合法、边界圆拟合法和最小二乘法拟合求出目标的精确位置。理论分析和实验表明 ,完整圆目标时 ,测量精度可达 1/4像素 ;而当目标图像有缺陷时 ,边界圆拟合法和最小二乘法拟合测量误差增加较少 ,边界重心法的测量误差增加较多。在具体情况下 ,算法及参数应该作相应的调整

One method of super-resolution position measurement for uniform circle objects by video images

One method of super resolution position measurement for uniform circle objects by video images is suggested The principle of super resolution position measurement is introduced, the scheme and test are finished, and the test results are satisfiable First, when objects are still, a serial video images are catched These images are added and averaged in order to decrease the noise Second, the object boundarys are obtained by Roberts edge operator, and the approximate positions of objects are obtained Last,the exact positions of objects are obtained by method of boundary weight center fitting, circle curve fitting and minimun mean square error fitting The analyse and test results show that the measurement error can be less than 1/4 pixel for full objects; for defective objects, the error increases few for using circle curve fitting and minimun mean square error fitting, and increases more for using boundary weight center fitting For different instance,the fitting arithmetic and parameter should be modulated