摘要
建立了基于计算机视觉的图像拼接测量系统,采用基于摄像机径向畸变的非线性优化模型,同时确定图像中心点,运用空间点重建的方法,建立起板类件特征的二维平面坐标与空间坐标的联系。采用旋转矩阵和平移矩阵的方法使分块测量数据统一到同一个视场下,最后匀化、拼接成为一个整体。最后通过几种不同形状的典型件验证了此种方法的合理性。
Measurement system of image integration based on computer vision was constructed. Non-linear optimizing model of camera radial aberrance was adopted At the same time image center points were confirmed, and with the method of space point construction the relationship of from 2D to 3D in sheet metal part was erected. Based on the method of rotation and movement matrices dividing region survey data was united into the same field, it was smoothed and integrated to the whole. At last the rationality of this method was tested by several different representative work pieces.
出处
《塑性工程学报》
CAS
CSCD
2003年第5期30-34,共5页
Journal of Plasticity Engineering
基金
国家"十五"重点科技攻关资助项目(2001BA203B11)
国家自然基金(50275063)
关键词
立体视觉
图像拼接
匹配
空间重建
仿射变换
stereo vision
image integration
matching
space reconstruction
affine transformation
