Image-based 3D modeling has always been a hot topic in the Held of computer vision. Currently image-based modeling method mostly use algorithm to pick feature points automatically, which often could not depict the contour accurately. To deal with this problem, an interactive modeling method based on video was proposed. Based on CAHV camera model, manually pick key feature points in any image to fix on the contour, calculate the depth with the camera parameters and partial information of automatic picked feature points,and use texture mapping obtain the 3D model in the end. The experimental results show the method can accurately extract the contour, obtain 3 D model with high degree of realism, and is effective for image-based modeling.%基于图像的三维建模技术一直是计算机视觉领域的研究热点.目前的基于图像的建模方法多使用算法自动提取图像特征点,常常无法准确描述场景物体轮廓,针对此问题提出了一种基于视频的交互式建模方法.基于CAHV摄像机模型,在任意图像中手动选取关键特征点确定物体轮廓,并结合摄像机参数及部分自动提取的特征点信息计算深度,最后通过纹理映射技术获得三维模型.实验证明,该方法可以准确提取物体轮廓,建模结果真实感强,是一种有效的基于图像的建模方法.
展开▼
