论文部分内容阅读
在层去图象法测量系统中，物体的空间坐标与截面图象坐标之间存在着复杂的非线性映射关系。如果采用完全理想条件和线性几何失真方法来标定系统，则会影响测量精度。本文提出了一种基于神经网络的标定方法，显著地提高了测量系统的精度。
In the layer to image measurement system, there is a complex nonlinear mapping between the object's space coordinates and the cross-sectional image coordinates. Accuracy of measurement can be affected if the system is calibrated using perfectly ideal conditions and linear geometric distortion methods. In this paper, a calibration method based on neural network is proposed, which improves the accuracy of measurement system significantly.