摘要
基于双目的立体视觉方法,介绍了一套由双工业CCD构成的立体视觉系统,设计了一套切实可行的三维重建系统方案,其中包括图像获取模块,基于OpenCV的摄像机标定系统,SIFT算法实现特征点提取与立体匹配,深度信息计算,OpenGL三维模型重构几大模块。系统各模块经过试验测试和验证,能够通过两幅图像恢复出物体三维可见表面几何形状,充分发挥了OpenCV函数库的功能,基本上能满足三维重建目标的要求,尤其对城市景观的三维重建有较大应用价值。
Based on binocular stereo vision methods, introduced a three - dimensional visual system constituted by a pair of industrial CCD,designed a practical system of three- dimensional programme, including image acquisition module, camera calibration system based on the OpenCV, feature extraction and matching based on SIFT algorithm, in - depth information calculated, 3D model reconstruction based OpenGL several modules. After the test and verify, it can restoration 3D visible surface geometry of the object through two images, give full play to the OpenCV library functions, it can basically meet ,the requirements of three - dimensional reconstruction objectives, in particular has a greater value in the urban landscape reconstruction.
出处
《计算机技术与发展》
2009年第6期127-130,共4页
Computer Technology and Development
基金
国家自然科学基金(10676029)
