A layered stereo matching algorithm using image segmentation and global visibility constraints
Citations
3,302 citations
Additional excerpts
...Ç...
[...]
969 citations
Cites background or methods from "A layered stereo matching algorithm..."
...Stereo matching continues to be an active research area as is proven by a large number of recent publications dedicated to this topic [1, 2, 4, 6, 9, 12]....
[...]
...Recently, segment-based methods [1, 2, 4, 6, 11] have attracted attention due to their good performance....
[...]
...Fourth, an optimal disparity plane assignment (optimal labeling) is approximated using greedy [1, 11] or graph cuts [2, 4, 6] optimization....
[...]
822 citations
Cites background from "A layered stereo matching algorithm..."
...They have a long history, dating back to [2] and were shown to be very successful on the Middleburry benchmark [22, 15, 3, 24] as well as on KITTI [25, 26, 27]....
[...]
687 citations
Cites methods from "A layered stereo matching algorithm..."
..., [2, 12]) extract several planes using an initial disparity map in the first step....
[...]
420 citations
Cites background from "A layered stereo matching algorithm..."
...Usually, depth estimation algorithms attempt to match corresponding signal components in two or more original cameras, using a matching function [44] with different area support and size [ 3 ]....
[...]
References
13,200 citations
"A layered stereo matching algorithm..." refers methods in this paper
...We calibrated the cameras using the method described in Zhang (2000) and transformed the images into epipolar geometry....
[...]
...A similar approach was taken by Zhang and Kambhamettu (2002). We follow their idea to measure the reliability of a segment’s disparity information by the density of valid points....
[...]
7,458 citations
"A layered stereo matching algorithm..." refers background or methods in this paper
...The reader is referred to Scharstein and Szeliski (2002) for an extensive survey of prior work....
[...]
...…the graph-based method of Kolmogorov and Zabih (2002), a belief propagation algorithm (Sun et al., 2003), an algorithm using dynamic programming (Birchfield and Tomasi, 1999b) and an implementation of sum-ofsquared-differences (SSD) by Scharstein and Szeliski (2002) that uses shiftable windows....
[...]
...We demonstrated the performance of the proposed algorithm using the test bed of Scharstein and Szeliski (2002)....
[...]
...We evaluated our algorithm using the test bed proposed by Scharstein and Szeliski (2002)....
[...]
...For quantitative evaluation, Scharstein and Szeliski (2002) measure the percentage of unoccluded pixels whose absolute disparity error is greater than one....
[...]
7,413 citations
3,199 citations
"A layered stereo matching algorithm..." refers methods in this paper
...Boykov et al. (2001) present an efficient greedy algorithm based on graph cuts....
[...]
...(Hirschmüller et al., 2002; Mühlmann et al., 2002). In our work, too, we take advantage of the efficient incremental computation of the DSI in the generation of the initial disparity map. Furthermore, we use different window sizes starting with smaller windows in order to preserve fine image details wherever possible. Disparity estimates for less-textured regions are then derived by using larger window sizes. Cooperative approaches (Zitnick and Kanade, 2000; Zhang and Kambhamettu, 2002; Mayer, 2003) locally compute matching scores using match windows. Nevertheless, they show “global behaviour” by iteratively refining the correlation scores using the uniqueness and continuity constraints. Zhang and Kambhamettu (2002) take advantage of image segmentation in the calculation of the initial matching scores. Furthermore, they exploit the results of the segmentation in their choice of local support area, preventing the support area from overlapping a depth discontinuity. Similarly to Zhang and Kambhamettu (2002), we use the output of image segmentation to propagate reliable disparity information inside a segment....
[...]
...(Hirschmüller et al., 2002; Mühlmann et al., 2002). In our work, too, we take advantage of the efficient incremental computation of the DSI in the generation of the initial disparity map. Furthermore, we use different window sizes starting with smaller windows in order to preserve fine image details wherever possible. Disparity estimates for less-textured regions are then derived by using larger window sizes. Cooperative approaches (Zitnick and Kanade, 2000; Zhang and Kambhamettu, 2002; Mayer, 2003) locally compute matching scores using match windows. Nevertheless, they show “global behaviour” by iteratively refining the correlation scores using the uniqueness and continuity constraints. Zhang and Kambhamettu (2002) take advantage of image segmentation in the calculation of the initial matching scores....
[...]
1,840 citations
"A layered stereo matching algorithm..." refers methods in this paper
...We further evaluated the proposed algorithm on a more complex scene using the Teddy test set taken from Scharstein and Szeliski (2003)....
[...]
...Furthermore, we present disparity maps for a more complex scene that was taken from Scharstein and Szeliski (2003) and for a self-recorded stereo pair....
[...]
...Furthermore, we applied our method to a more complex image pair taken from Scharstein and Szeliski (2003) and to self-recorded data....
[...]
...16c. Pixels for which the method of Scharstein and Szeliski (2003) fails to produce the ground truth are coloured black....
[...]