Filters








3,267 Hits in 3.8 sec

Displacement-Invariant Matching Cost Learning for Accurate Optical Flow Estimation [article]

Jianyuan Wang, Yiran Zhong, Yuchao Dai, Kaihao Zhang, Pan Ji, Hongdong Li
2020 arXiv   pre-print
Our key innovation is to decouple the connection between 2D displacements and learn the matching costs at each 2D displacement hypothesis independently, ie, displacement-invariant cost learning.  ...  Learning matching costs has been shown to be critical to the success of the state-of-the-art deep stereo matching methods, in which 3D convolutions are applied on a 4D feature volume to learn a 3D cost  ...  Moreover, our method does not need to construct a 5D volume and can efficiently compute the matching cost by 2D convolutions.  ... 
arXiv:2010.14851v1 fatcat:g5vbz2ns4nbxrbvr6rp5g2bjtm

Volumetric Correspondence Networks for Optical Flow

Gengshan Yang, Deva Ramanan
2019 Neural Information Processing Systems  
Well-known techniques for doing so make use of a cost volume, typically a 4D tensor of match costs between all pixels in a 2D image and their potential matches in a 2D search window.  ...  Stateof-the-art (SOTA) deep networks for flow/stereo make use of such volumetric representations as internal layers.  ...  Acknowledgements: This work was supported by the CMU Argo AI Center for Autonomous Vehicle Research.  ... 
dblp:conf/nips/YangR19 fatcat:aud36plzebfa5avg7xnswmvxwi

Iterated dynamic programming and quadtree subregioning for fast stereo matching

Carlos Leung, Ben Appleton, Changming Sun
2008 Image and Vision Computing  
A quadtree subregioning process is also used for efficient computation of a matching cost volume where iterated dynamic programming operates on.  ...  The application of energy minimisation methods for stereo matching has been demonstrated to produce high quality disparity maps.  ...  For fast stereo matching, efficient computation of a matching cost volume is essential.  ... 
doi:10.1016/j.imavis.2007.11.013 fatcat:b2377iukqbfc3le4bdrxjik4bi

Blur and Contrast Invariant Fast Stereo Matching [chapter]

Matteo Pedone, Janne Heikkilä
2008 Lecture Notes in Computer Science  
The descriptors are based on local-phase quantization, they can be computed very efficiently and encoded in a limited number of bits.  ...  We propose a novel approach for estimating a depth-map from a pair of rectified stereo images degraded by blur and contrast change.  ...  Blur Robust Stereo Matching In this section we present blur and contrast invariant descriptors based on quantized local phase.  ... 
doi:10.1007/978-3-540-88458-3_80 fatcat:nrqawy2ag5f2hc6jnyrjhyplpi

Dense Segmentation-Aware Descriptors [chapter]

Eduard Trulls, Iasonas Kokkinos, Alberto Sanfeliu, Francesc Moreno-Noguer
2016 Dense Image Correspondences for Computer Vision  
We integrate this idea with Dense SIFT, and also with Dense Scale and Rotation Invariant Descriptors (SID), delivering descriptors that are densely computable, invariant to scaling and rotation, and robust  ...  We apply our approach to standard benchmarks on large displacement motion estimation using SIFT-flow and widebaseline stereo, systematically demonstrating that the introduction of segmentation yields clear  ...  The computational cost of matching two images with the SIFT-flow framework depends on the size of the descriptors, varying from ∼14 seconds for SIFT (the smallest) to ∼80 seconds for SID/SSID, and ∼10  ... 
doi:10.1007/978-3-319-23048-1_5 fatcat:7fojz5b6z5fuhpys742vbpl35m

Dense Segmentation-Aware Descriptors

Eduard Trulls, Iasonas Kokkinos, Alberto Sanfeliu, Francesc Moreno-Noguer
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
We integrate this idea with Dense SIFT, and also with Dense Scale and Rotation Invariant Descriptors (SID), delivering descriptors that are densely computable, invariant to scaling and rotation, and robust  ...  We apply our approach to standard benchmarks on large displacement motion estimation using SIFT-flow and widebaseline stereo, systematically demonstrating that the introduction of segmentation yields clear  ...  The computational cost of matching two images with the SIFT-flow framework depends on the size of the descriptors, varying from ∼14 seconds for SIFT (the smallest) to ∼80 seconds for SID/SSID, and ∼10  ... 
doi:10.1109/cvpr.2013.372 dblp:conf/cvpr/TrullsKSM13 fatcat:sc7wqaiqxfd3lhhtca44see6vi

Entropy based Log Chromaticity Projection for Real-time Stereo Matching

U. Raghavendra, Makkithaya Krishnamoorthi, A.K. Karunakar
2012 Procedia Technology - Elsevier  
Most of the existing stereo matching algorithms will assume a similar corresponding color values between stereo images.  ...  This paper proposes an entropy minimization based log chromaticity projection for stereo image, thereby extracting the invariant image, which is independent of illumination and color.  ...  The generated invariant image is independent of lighting and it will be a better cost function for matching the stereo images.  ... 
doi:10.1016/j.protcy.2012.10.027 fatcat:jxft4tk54fgqvlky2oaqxo6nh4

An Optimized Mean Shift Filtering Technique to Image Representation Through Disparity Map for Large Scale Stereo Images

Kavitha, Balakrishnan
2017 Application and Theory of Computer Technology  
For pixels corresponding to different depths, an adaptive iterative algorithm is proposed to choose optimal frames for stereo matching, which can take advantage of continuously pose-changing imaging and  ...  In order to overcome these issues, in this paper, we propose a stereo matching algorithm based on the optimized mean shift image filtering to compute the depth estimation in the 3D information's along  ...  Using linear interpolation, reconstruction of images can be given clearly for stereo images. Use of multiple fitting produces a smooth disparity surface along less computational cost.  ... 
doi:10.22496/atct20170104126 fatcat:zqokxveadnerjgcj32savjgcxa

A variational framework for simultaneous motion and disparity estimation in a sequence of stereo images

Wided Miled, Beatrice Pesquet-Popescu, Wael Cherif
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
In order to reduce computational complexity and improve estimation accuracy, the two motion fields, for the left and right sequences, and the disparity field of the current stereo pair are jointly estimated  ...  In this paper, we present a variational framework for joint disparity and motion estimation in a sequence of stereo images.  ...  Joint estimation of disparity and motion displacement fields is an efficient way to benefit from these relationships, leading therefore to improving results while reducing the computational cost.  ... 
doi:10.1109/icassp.2009.4959690 dblp:conf/icassp/MiledPC09 fatcat:mny7fvdmvfdt3fnmmnhgp6coj4

ASV

Yu Feng, Paul Whatmough, Yuhao Zhu
2019 Proceedings of the 52nd Annual IEEE/ACM International Symposium on Microarchitecture - MICRO '52  
Firstly, we propose a new stereo algorithm, invariant-based stereo matching (ISM), that achieves significant speedup while retaining high accuracy.  ...  The key to ASV is to exploit unique characteristics inherent to stereo vision, and apply stereo-specific optimizations, both algorithmically and computationally. We make two contributions.  ...  Invariant-based Stereo Matching This section introduces our new invariant-based stereo matching algorithm (ISM).  ... 
doi:10.1145/3352460.3358253 dblp:conf/micro/FengW019 fatcat:kdsijuwfwzhsbgc2dkiglw4dl4

AdaStereo: A Simple and Efficient Approach for Adaptive Stereo Matching [article]

Xiao Song, Guorun Yang, Xinge Zhu, Hui Zhou, Zhe Wang, Jianping Shi
2021 arXiv   pre-print
Secondly, we design an efficient parameter-free cost normalization layer for internal feature-level alignment.  ...  Compared to previous methods for adaptive stereo matching, our AdaStereo realizes a more standard, complete and effective domain adaptation pipeline.  ...  Introduction Stereo matching is a fundamental problem in computer vision.  ... 
arXiv:2004.04627v3 fatcat:zlynhptskvcz3cex4ubsmekeaq

Know Your Limits: Accuracy of Long Range Stereoscopic Object Measurements in Practice [chapter]

Peter Pinggera, David Pfeiffer, Uwe Franke, Rudolf Mester
2014 Lecture Notes in Computer Science  
Evaluation benchmarks for stereo correspondence algorithms, such as the popular Middlebury and KITTI frameworks, provide important reference values regarding dense matching performance, but do not sufficiently  ...  We present a comprehensive statistical evaluation of selected state-of-the-art stereo matching approaches on an extensive dataset and establish reference values for the precision limits actually achievable  ...  Sub-pixel results are obtained by fractional sampling of the disparity space and/or a curve fit to the computed matching cost volume [30] .  ... 
doi:10.1007/978-3-319-10605-2_7 fatcat:rgzx3yjvajh27fbxvrfxbhmsny

Motion Estimations based on Invariant Moments for Frames Interpolation in Stereovision

Margarita Favorskaya, Dmitriy Pyankov, Aleksei Popov
2013 Procedia Computer Science  
A novel method improves the frames interpolation by forming an invariant set of local motion vectors. First, the motion in a scene is estimated by block-matching algorithm.  ...  Such approach provides smooth motion that significantly improves the resulting stereo video sequence. Experimental results show the efficiency of frames interpolation based on such approach.  ...  Second, an accurate motion in detected moving regions is determined by usage the invariant moments. The calculation of any moments is a high computer cost procedure.  ... 
doi:10.1016/j.procs.2013.09.196 fatcat:6idg6yge7bc7vin27v6vpsmt74

Automatic Georeferencing of Aerial Images Using Stereo High-Resolution Satellite Images

Jaehong Oh, Charles K. Toth, Dorota A. Grejner-Brzezinska
2011 Photogrammetric Engineering and Remote Sensing  
The matching between aerial and satellite stereo images is based on the SIFT (Scale-Invariant Feature Transform) features, and outliers are pruned utilizing RANSAC (RANdom SAmple Consensus).  ...  The use of stereo images can avoid the impact of relief displacement and requires no DSM to obtain ground heights.  ...  By stereo matching between the epipolar resampled images, 3D ground coordinates for each matching point can be computed.  ... 
doi:10.14358/pers.77.11.1157 fatcat:mgony7vmq5ealmpq36xeetggbu

DAISY Filter Flow: A Generalized Discrete Approach to Dense Correspondences

Hongsheng Yang, Wen-Yan Lin, Jiangbo Lu
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
Though significant advance has been made towards estimating dense stereo and optical flow fields for two images adjacent in viewpoint or in time, building reliable dense correspondence fields for two general  ...  Inspired by the recent PatchMatch Filter technique, we leverage and extend a few established methods: 1) DAISY descriptors, 2) filter-based efficient flow inference, and 3) the Patch-Match fast search.  ...  DFF yields dense coherent matches consistently. NRDC gives no match for different scenes. optical flow [23] and stereo matching [19] .  ... 
doi:10.1109/cvpr.2014.435 dblp:conf/cvpr/YangLL14a fatcat:m2whgc7ajjaezmlcxpyw73kaci
« Previous Showing results 1 — 15 out of 3,267 results