Large Scale Object Measurement Based on Data Fusion of Stereo Camera's Multi-viewpoint Images

Abstract

Multi-view photogrammetry is a method to obtain the spatial position of the target points in the field of view by fusing image data from multi-viewpoint, it's suitable for the measurement of large scale object's surface dimension parameters, which may under the circumstance of occlusion. The multi-view photogrammetry based on single camera needs to arrange scale datum into the measurement area, image acquisition process is complex, and requires a high overlap ratio between adjacent images. In this paper, a method of multi-view photogrammetry using stereo camera is presented, this method not only inherits the advantages of large measurement range in multi-view reconstruction, but also does not need to arrange scale datum, and the image acquisition process is simpler. Firstly the internal parameters and solid connection of two cameras installed on the fixed rod are obtained according to camera calibration. Then we use the stereo camera to acquire the images of the area to be measured and get the spatial point cloud by binocular intersection. Calculate the transformation relation of the coordinate system of point cloud in adjacent viewpoint by the method of pose estimation, and transform the camera's parameters and point cloud's coordinate of each moment to the specified coordinate system. The position and pose of the camera and the coordinate of point cloud are optimized by using bundle adjustment, the size parameters and deformation information can be calculated from the point cloud coordinates. The test results show that in the range of 5 meters the measurement's error is 3mm, the average error is 1mm. This method is suitable for the measurement of large scale object and scene, the algorithm is stable and reliable.