基于三维VR技术的小区域目标重建研究

摘要

针对当前重建目标区域缺失图像信息时局限性较大,提出一种结合三维VR技术的小区域目标重建方法.对小区域目标进行相机标定,在相机标定过程中,引入小区域目标二维图像物理参数组建相机成像模型,利用直接法求解小区域目标成像参数,结合相机标定的结果利用SURF算法对小区域目标二维图像提取特征点.寻找小区域目标图像内任意一个主方向,构成各个特征点的图像旋转不变量,以该不变量为依据利用归一化互相关方法计算出小区域目标两图像块之间的近似程度,通过对极几何的性质约束小区域目标特征匹配点计算基础矩阵,以获得的小区域目标投影矩阵和特征点的匹配关系为依据,完成小区域目标空间点的三维重建.实验结果表明该方法准确率及实用价值较高.%The current reconstruction of the target area has strong limitation while image information is missing,so a small-area target reconstruction method combining with 3D VR technology is proposed. The small-area target is performed with camera cali-bration,in which the 2D image′s physical parameters of small-area target are introduced to establish the camera imaging model, the direct method is used to solve the imaging parameters of small-area target,and the SURF algorithm is adopted to extract the feature points of 2D image of small-area target according to the camera calibration results. The arbitrary main direction within the small-area target image is found out to compose the image rotation invariant of each feature point. On this basis,the nor-malized cross-correlation method is used to calculate the approximation degree between two image blocks of small-area target. Ac-cording to the property of epipolar geometry,the feature matching points of small-area target is restrained to calculate the basis matrix. The matching relation between the obtained small-area target projection matrix and feature points is taken as the basis to realize the 3D reconstruction of small-area target spatial points. The experimental results show that the method has high accuracy and high practical value.