基于二维灰度图像的“非规则形体”三维重建

摘要

3D reconstruction from the 2D grey image has been a research hot spot in the field of computer vision.After making an intensive study on the hot spot,a method according to the characteristic points of the profile curves to complete the reconstruction of the irregular object is proposed.The method extracts the irregular object from the 2D grey image as the research target and adopts corner detection and the detection-threshold to acquire main characteristic points and subprime points.According to the main points to determine the matching relationship of the profile curves,thereby the segmentation-concatenation is completed,according to the subprime points and the shortest diagonal optimization strategy,the triangular plane-concatenation is finished,finally,the 3D reconstruction of irregular object surface is realized.The experiment indicates that the method could conveniently reconstruct the 3D shape of the irregular object according to its 2D grey image.%由二维灰度图像恢复形体的三维形状已经成为计算机视觉领域的一个研究热点,对该问题进行了深入的研究,提出了一种基于单幅图像,利用轮廓线特征点进行非规则形体三维重构的方法;该方法从一幅二维灰度图像中提取非规则形体作为研究对象,通过角点检测方法并结合检测阈值获取每条轮廓线的主次特征点;根据主特征点确定轮廓线间的形状匹配关系完成分段连接,根据次特征点并结合最小对角线优化策略完成三角面片拼接,最终实现非规则形体表面的三维重建;实验证明,该方法可以根据形体的二维灰度图像方便有效的重建出其三维形状。