论文部分内容阅读
立体视觉系统已广泛应用于三维形体检测等诸多方面 ,针对该系统物点到像点非线性关系的确定技术 ,提出了基于人工神经网络的标定、测量方法 ,同时提出了加快网络训练速度、提高网络训练精度的新方法 ,并给出了学习和测量结果 ,验证了此方法的正确性。这种方法降低了对系统本身精度的要求 ,且不用考虑透镜畸变的影响。是对立体视觉系统数学建模的有益探索 ,具有一定的理论意义和现实意义。
Stereo vision system has been widely used in three-dimensional shape detection and many other aspects. According to the determination of the non-linear relationship between the point and the point of the system, this paper proposes a method of calibration and measurement based on artificial neural network. At the same time, it proposes to speed up the network training and improve Network training accuracy of the new method, and gives the learning and measurement results verify the correctness of this method. This method reduces the accuracy of the system itself, without regard to the effects of lens distortion. It is a useful exploration of mathematical modeling of stereo vision system, which has a certain theoretical and practical significance.