Analysis of region based stereo matching algorithms
01 Dec 2014-pp 1-4
TL;DR: Various region based algorithms for stereo vision used for generating disparity map are implemented and their results are compared on the basis of window size and time complexity.
Abstract: Stereo Vision has created an impact on research in the area of computer vision system. It has many applications like 3D scene reconstruction, mobile robotics to detect obstacles, surgical robotics, etc. This paper focuses on various region based algorithms for stereo vision. They are used for generating disparity map. This disparity map is used to find out the distance of the objects in the images from camera. The researchers have implemented these algorithms and their results are compared on the basis of window size and time complexity.
Citations
More filters
Posted Content•
TL;DR: In this paper, the authors proposed Adaptive Deconvolution-based disparity matching Net (ADSM net) by adding deconvolution layers to learn how to enlarge the size of input feature map for the following convolution layers.
Abstract: In deep learning-based local stereo matching methods, larger image patches usually bring better stereo matching accuracy. However, it is unrealistic to increase the size of the image patch size without restriction. Arbitrarily extending the patch size will change the local stereo matching method into the global stereo matching method, and the matching accuracy will be saturated. We simplified the existing Siamese convolutional network by reducing the number of network parameters and propose an efficient CNN based structure, namely Adaptive Deconvolution-based disparity matching Net (ADSM net) by adding deconvolution layers to learn how to enlarge the size of input feature map for the following convolution layers. Experimental results on the KITTI 2012 and 2015 datasets demonstrate that the proposed method can achieve a good trade-off between accuracy and complexity.
2 citations
12 Jun 2019
TL;DR: The experimental results show that for image pairs with a resolution of $640\times 480$, the final processing speed of the algorithm on the FPGA can reach 32fps, which can meet the general real-time requirements.
Abstract: A census stereo matching algorithm with adaptive transform window is presented in this paper, the hardware architecture of the algorithm is designed, the RTL codes of the architecture are simulated and synthesized to Altera’s FPGA, the system is tested on FPGA development board. The experimental results show that for image pairs with a resolution of $640\times 480$, the final processing speed of the algorithm on the FPGA can reach 32fps, which can meet the general real-time requirements.
1 citations
Cites methods from "Analysis of region based stereo mat..."
...Stereo matching is the most important part of binocular stereo vision, the basic idea of stereo matching is to capture the images from the binocular camera, with the left image as the reference image and the right image as the target image, and then obtained the disparity map by the stereo matching algorithm [1]....
[...]
References
More filters
TL;DR: A new methodology for reliably solving the correspondence problem between sparse sets of points of two or more images is proposed, which performs correspondence and outlier rejection in a single step and achieves global optimality with feasible computation.
Abstract: We propose a new methodology for reliably solving the correspondence problem between sparse sets of points of two or more images. This is a key step inmost problems of computer vision and, so far, no general method exists to solve it. Our methodology is able to handle most of the commonly used assumptions in a unique formulation, independent of the domain of application and type of features. It performs correspondence and outlier rejection in a single step and achieves global optimality with feasible computation. Feature selection and correspondence are first formulated as an integer optimization problem. This is a blunt formulation, which considers the whole combinatorial space of possible point selections and correspondences. To find its global optimal solution, we build a concave objective function and relax the search domain into its convex-hull. The special structure of this extended problem assures its equivalence to the original one, but it can be optimally solved by efficient algorithms that avoid combinatorial search. This methodology can use any criterion provided it can be translated into cost functions with continuous second derivatives.
226 citations
"Analysis of region based stereo mat..." refers background in this paper
...This criterion provides the simplest formulation of the correspondence problem so it can be used to match large numbers of features [7]....
[...]
01 Dec 2010
TL;DR: An obstacle detection algorithm based on U-V disparity map analysis is presented, and the experimental results demonstrate that this algorithm is efficient and has the strong robustness to the change of illumination and obstacles.
Abstract: An obstacle detection algorithm based on U-V disparity map analysis is presented in this paper We firstly introduce the obstacle detection mechanism using U-V disparity map generated from stereo match disparity map, and then an obstacle detection algorithm based on analysis of U-V disparity map characteristic is described The combination of straight line fitting and the standard Hough Transform are used in processing the initial disparity map, and the experimental results demonstrate that this algorithm is efficient and has the strong robustness to the change of illumination and obstacles
16 citations
"Analysis of region based stereo mat..." refers methods in this paper
...The U-V disparity map is used to characterize road surface, road structure and to detect the obstacle [3]....
[...]
31 Aug 2011
TL;DR: The proposed method is able to classify the scene of further stereo image pairs as traversable or non-traversable, which is often the first step towards more advanced autonomous robot navigation behaviours.
Abstract: This work presents a machine learning method for terrain's traversability classification. Stereo vision is used to provide the depth map of the scene. Then, a v-disparity image calculation and processing step extracts suitable features about the scene's characteristics. The resulting data are used as input for the training of a support vector machine (SVM). The evaluation of the traversability classification is performed with a leave-one-out cross validation procedure applied on a test image data set. This data set includes manually labeled traversable and nontraversable scenes. The proposed method is able to classify the scene of further stereo image pairs as traversable or non-traversable, which is often the first step towards more advanced autonomous robot navigation behaviours.
14 citations
"Analysis of region based stereo mat..." refers methods in this paper
...SAD algorithm is used to take decision about the obstacle [2]....
[...]
09 Dec 2001
TL;DR: A new feature based algorithm for stereo correspondence that detects and matches dense features between the left and right images of a stereo pair, producing a semi-dense disparity map.
Abstract: We present a new feature based algorithm for stereo correspondence. Most of the previous feature based methods match sparse features like edge pixels, producing only sparse disparity maps. Our algorithm detects and matches dense features between the left and right images of a stereo pair, producing a semi-dense disparity map. Our dense feature is defined with respect to both images of a stereo pair, and it is computed during the stereo matching process, not a preprocessing step. In essence, a dense feature is a connected set of pixels in the left image and a corresponding set of pixels in the right image such that the intensity edges on the boundary of these sets are stronger than their matching error (which is basically the difference in intensities between corresponding boundary pixels). Our algorithm produces accurate semi-dense disparity maps, leaving featureless regions in the scene unmatched. It is robust, requires little parameter tuning, can handle brightness differences between images, and is fast (linear complexity).
12 citations
"Analysis of region based stereo mat..." refers background in this paper
...Segmentation based area matching can also be implemented [1]....
[...]
01 Jan 2013
TL;DR: Quantitative results show that the method of reconstructing disparity maps from sub-sampled data is only marginally inferior to the full disparity map computed by correlation based method.
Abstract: In this paper we propose a new technique for computing disparity maps. This paper is based on the popular correlation based disparity map computation approach. The task is to estimate a dense disparity map given two stereo images. Instead of computing the full disparity map, we only compute the disparity values for randomly selected epipolar lines. The output of this exercise is a sampled disparity map; from the sampled disparity map, the full map is reconstructed by exploiting the sparsity of the map in Fourier domain using Compressed Sensing techniques. Quantitative results show that our method of reconstructing disparity maps from sub-sampled data is only marginally inferior to the full disparity map computed by correlation based method.
4 citations
"Analysis of region based stereo mat..." refers methods in this paper
...The full disparity map is computed using correlation based method [6]....
[...]