摘要
为了解决摄像机标定存在的若干问题 ,根据立体视觉原理 ,提出了基于神经网络的双目视觉系统标定方法。通过对双目摄像机的有效视场分析 ,确定了一次测量面积 ,并把像对视差作为网络输入 ,建立空间点世界坐标与图像坐标非线性映射关系 ,使系统不经过复杂的摄像机内外参数标定 ,就能直接提取物体的三维信息 ,增加了系统的灵活性。实验证明 。
On the basis of stereo vision, a method for visions system calibration based on neural network is proposed. The once measurement area is calculated after the effective view field of binocular camera is analyzed. Not only 3D information of the object is directly acquired, but also the relation between spatial points and image points is established by training the network without the parameters of the cameras calibrated, while the disparity of image pairs is input into the network. Experiment results with panel curve surface are given to verify the method.
出处
《光学技术》
EI
CAS
CSCD
2001年第4期302-304,共3页
Optical Technique
基金
国家"8 63计划"CIMS项目资助
