Analysis of KITTI data for stereo analysis with stereo confidence measures


by R Haeusler, R Klette
Abstract:
The recently published KITTI stereo dataset provides a new quality of stereo imagery with partial ground truth for benchmarking stereo matchers. Our aim is to test the value of stereo confidence measures (e.g. a left-right consistency check of disparity maps, or an analysis of the slope of a local interpolation of the cost function at the taken minimum) when applied to recorded datasets, such as published with KITTI. We choose popular measures as available in the stereo-analysis literature, and discuss a naive combination of these. Evaluations are carried out using a sparsification strategy. While the best single confidence measure proved to be the right-left consistency check for high disparity map densities, the best overall performance is achieved with the proposed naive measure combination. We argue that there is still demand for more challenging datasets and more comprehensive ground truth. © 2012 Springer-Verlag.
Reference:
Analysis of KITTI data for stereo analysis with stereo confidence measures (R Haeusler, R Klette), In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 7584 LNCS, 2012.
Bibtex Entry:
@inproceedings{haeusler2012analysismeasures,
author = "Haeusler, R and Klette, R",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "158--167",
title = "Analysis of KITTI data for stereo analysis with stereo confidence measures",
volume = "7584 LNCS",
year = "2012",
abstract = "The recently published KITTI stereo dataset provides a new quality of stereo imagery with partial ground truth for benchmarking stereo matchers. Our aim is to test the value of stereo confidence measures (e.g. a left-right consistency check of disparity maps, or an analysis of the slope of a local interpolation of the cost function at the taken minimum) when applied to recorded datasets, such as published with KITTI. We choose popular measures as available in the stereo-analysis literature, and discuss a naive combination of these. Evaluations are carried out using a sparsification strategy. While the best single confidence measure proved to be the right-left consistency check for high disparity map densities, the best overall performance is achieved with the proposed naive measure combination. We argue that there is still demand for more challenging datasets and more comprehensive ground truth. © 2012 Springer-Verlag.",
doi = "10.1007/978-3-642-33868-7_16",
isbn = "9783642338670",
issn = "0302-9743",
eissn = "1611-3349",
issue = "PART 2",
language = "eng",
}