Shape-from-Silhouettes-master

所属分类:图形图像处理
开发工具:matlab
文件大小:5923KB
下载次数:0
上传日期:2020-08-06 11:32:43
上 传 者袁大头2011
说明:  基于 shape of silhouettes的三维重建 matlab程序
(# Shape-from-Silhouettes 3D reconstruction with shape of silhouettes ## Discription The task is to reconstruct 3D object from multiple calibrated images by implementing a naive silhouette extraction algorithm. Images and camera calibrations are provided.)

文件列表:
code_exercise\.svn\all-wcprops (240, 2018-02-02)
code_exercise\.svn\entries (364, 2018-02-02)
code_exercise\.svn\text-base\exercise7.m.svn-base (3434, 2018-02-02)
code_exercise\3Dmodel.fig (698187, 2018-02-02)
code_exercise\exercise7.m (4227, 2018-02-02)
data\.svn\all-wcprops (4563, 2018-02-02)
data\.svn\entries (5838, 2018-02-02)
data\.svn\prop-base\david_00.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_01.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_02.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_03.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_04.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_05.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_06.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_07.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_08.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_09.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_10.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_11.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_12.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_13.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_14.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_15.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_16.jpg.svn-base (53, 2018-02-02)
data\.svn\prop-base\david_17.jpg.svn-base (53, 2018-02-02)
data\.svn\text-base\david_00.jpg.svn-base (109641, 2018-02-02)
data\.svn\text-base\david_00.pa.svn-base (179, 2018-02-02)
data\.svn\text-base\david_01.jpg.svn-base (108246, 2018-02-02)
data\.svn\text-base\david_01.pa.svn-base (180, 2018-02-02)
data\.svn\text-base\david_02.jpg.svn-base (109413, 2018-02-02)
data\.svn\text-base\david_02.pa.svn-base (180, 2018-02-02)
data\.svn\text-base\david_03.jpg.svn-base (108255, 2018-02-02)
data\.svn\text-base\david_03.pa.svn-base (180, 2018-02-02)
data\.svn\text-base\david_04.jpg.svn-base (105147, 2018-02-02)
data\.svn\text-base\david_04.pa.svn-base (181, 2018-02-02)
data\.svn\text-base\david_05.jpg.svn-base (104433, 2018-02-02)
data\.svn\text-base\david_05.pa.svn-base (179, 2018-02-02)
data\.svn\text-base\david_06.jpg.svn-base (107147, 2018-02-02)
data\.svn\text-base\david_06.pa.svn-base (179, 2018-02-02)
data\.svn\text-base\david_07.jpg.svn-base (104926, 2018-02-02)
... ...

# Shape-from-Silhouettes 3D reconstruction with shape of silhouettes ## Discription The task is to reconstruct 3D object from multiple calibrated images by implementing a naive silhouette extraction algorithm. Images and camera calibrations are provided. ### 1. Silhouette extraction Extract silhouettes from the provided images, using a simple thresholding technique. Adjust the silhoutteThreshold so that the silhouettes of the statue are clearly extracted. ![image](https://user-images.githubusercontent.com/2938***92/35713540-296a6336-07c8-11e8-8c35-4a137b701128.png) ### 2. Volume of interest Define the volume of interest. At first use a larger bounding box than necessary, just to make sure the statue has been included. Then make the bounding box tight to get better resolution. Similarly, we can start with a coarse grid, say 10 * 10 * 20, and then try a larger grid, at least *** * *** * 128. ### 3. Visual hull Compute the occupancy score at each voxel. For each voxel, transform the point from volume to world coordinates using the transformation provided. Then project the points into the image. If a projected point falls within the silhouette, add 1 to the score.

近期下载者

相关文件


收藏者