摘要
提出了一种基于平面的逆向Warping算法,用于根据多幅参考图像生成任意视点下的新图像.首先通过参考图像的深度信息来重建三维平面,然后寻找这些重建平面间的对应关系,并比较它们对空间平面同一部分的采样密度,以获得最好的采样结果.在生成新视点图像时,首先对那些采样密度最好的重建平面进行可见性判断,然后将其投影到新视点下,在此基础上求得目标图像上各点的深度,最后将目标图像上的点逆向Warping到相应的参考图像中,以获取它们的颜色值.同时,对于参考图像中不能被重建成平面的像素点,用正向Warping的方法对其进行处理.实验数据显示,此算法的时间效率不会随着参考图像的增加而急剧下降,并且由于只保存了参考图像中采样密度最好的重建平面,所以算法对存储量的要求也很小.
In this paper, a plane-based backward warping algorithm is proposed to generate novel views from multiple reference images. First the depth information is employed to reconstruct space planes from individual reference images and the potential occluding relationship between these planes is calculated. Then the planes that represent an identical space plane from different reference images are compared with each other to decide one with best sample rate preserved and used in the later warping period while the other samples are abandoned. While the image of a novel view is produced, the planes that pass the visibility test are projected onto the desired image from which people can get knowledge on the plane that the desired image pixels are warped from. Finally, pixels?depth of the desired image is calculated and then a backward warping is performed from these pixels to the reference images to obtain their colors. It is shown that the storage requirement and executing time of the proposed algorithm increases slowly with the number of reference images.
出处
《软件学报》
EI
CSCD
北大核心
2002年第7期1242-1249,共8页
Journal of Software
基金
国家自然科学基金资助项目(69873044)