【摘要】 摄像机标定技术是从二维平面图像中获取三维信息的重要前提,是计算机视觉的关键技术,在视觉导航、运动分析、三维重建等领域得到了广泛的研究和应用。文中针对多视点平行阵列摄像机系统实现了一种利用HALCON提供的丰富函数库,开发了八视点平行阵列摄像机标定。然后通过Maya构造虚拟实验场景,验证了所提方案的有效性。最后将所开发的程序移植到通用Windows平台,并集成视频采集系统中。更多还原

【Abstract】 Camera calibration is an important step of recovering 3D shape from 2D images and a key technology of computer vision. It has widely used in many fields. This paper proposes a marked-template based calibration method for multi viewpoint parallel camera array system. The method uses the marked-template and function library provided by HALCON to develop an eight viewpoint camera calibration process. Then it constructs a virtual experiment scene through Maya software to verify the method validity. Finally,the program is transplanted to the general platform under Windows OS,and integrated into the multi-viewpoint video acquisition system.更多还原