摘要
为了减小视觉坐标测量系统中特征点圆心像面投影坐标计算的误差,以改善系统的测量精度,提出了一种利用透视投影变换的直线不变性对其进行修正的新型光靶标。传统的误差减小方法是提高特征点的像边缘检测的精度,或者对特征点像面轮廓椭圆拟合的数学模型进行修正,算法复杂、适应性较差,且对精度的改善也达到了瓶颈。仿真试验表明,基于直线不变性的新型光靶标支持特征点-对应像面位置自动识别,较有效地修正了特征点圆心像面投影坐标计算的误差。
In order to reduce the calculation error of the coordinates of the control points＇ center in 3D vision coordinate measuring system so as to improve the testing accuracy of the system, a self-rectification light target based on projective invariance of lines using perspective transformation method is introduced. The Traditional methods to control the error are improving the edge detection precision of the control points or revising the mathematical model of the ellipse fitting for the edge of the image of the control points. These traditional methods are complicate and with poor adaptability, meanwhile these methods are poor on the improvement capacity of accuracy. Computer simulation indicated that this new light target supported the auto recognition of the control points and their corresponding projections on the image plane. Also, the error of the obtained coordinates of the control points＇ center under the image plane coordinate system was reduced.
出处
《新技术新工艺》
2010年第1期12-14,共3页
New Technology & New Process
关键词
光靶标
特征点
像面坐标系
自动识别
Light target, Control point, Image plane coordinate system, Auto recognition
