scholarly journals A Low-complexity End-to-end Stereo Matching Pipeline from Raw Bayer Pattern Images to Disparity Maps

IEEE Access ◽  
2021 ◽  
pp. 1-1
Author(s):  
Shengyu Gao ◽  
Hongyu Wang ◽  
Xin Lou
2021 ◽  
Vol 13 (2) ◽  
pp. 274
Author(s):  
Guobiao Yao ◽  
Alper Yilmaz ◽  
Li Zhang ◽  
Fei Meng ◽  
Haibin Ai ◽  
...  

The available stereo matching algorithms produce large number of false positive matches or only produce a few true-positives across oblique stereo images with large baseline. This undesired result happens due to the complex perspective deformation and radiometric distortion across the images. To address this problem, we propose a novel affine invariant feature matching algorithm with subpixel accuracy based on an end-to-end convolutional neural network (CNN). In our method, we adopt and modify a Hessian affine network, which we refer to as IHesAffNet, to obtain affine invariant Hessian regions using deep learning framework. To improve the correlation between corresponding features, we introduce an empirical weighted loss function (EWLF) based on the negative samples using K nearest neighbors, and then generate deep learning-based descriptors with high discrimination that is realized with our multiple hard network structure (MTHardNets). Following this step, the conjugate features are produced by using the Euclidean distance ratio as the matching metric, and the accuracy of matches are optimized through the deep learning transform based least square matching (DLT-LSM). Finally, experiments on Large baseline oblique stereo images acquired by ground close-range and unmanned aerial vehicle (UAV) verify the effectiveness of the proposed approach, and comprehensive comparisons demonstrate that our matching algorithm outperforms the state-of-art methods in terms of accuracy, distribution and correct ratio. The main contributions of this article are: (i) our proposed MTHardNets can generate high quality descriptors; and (ii) the IHesAffNet can produce substantial affine invariant corresponding features with reliable transform parameters.


Author(s):  
Fangrui Wu ◽  
Menglong Yang

Recent end-to-end CNN-based stereo matching algorithms obtain disparities through regression from a cost volume, which is formed by concatenating the features of stereo pairs. Some downsampling steps are often embedded in constructing cost volume for global information aggregation and computational efficiency. However, many edge details are hard to recover due to the imprudent upsampling process and ambiguous boundary predictions. To tackle this problem without training another edge prediction sub-network, we developed a novel tightly-coupled edge refinement pipeline composed of two modules. The first module implements a gentle upsampling process by a cascaded cost volume filtering method, aggregating global information without losing many details. On this basis, the second module concentrates on generating a disparity residual map for boundary pixels by sub-pixel disparity consistency check, to further recover the edge details. The experimental results on public datasets demonstrate the effectiveness of the proposed method.


2020 ◽  
Vol 64 (2) ◽  
pp. 20505-1-20505-12
Author(s):  
Hui-Yu Huang ◽  
Zhe-Hao Liu

Abstract A stereo matching algorithm is used to find the best match between a pair of images. To compute the cost of the matching points from the sequence of images, the disparity maps from video streams are estimated. However, the estimated disparity sequences may cause undesirable flickering errors. These errors result in low visibility of the synthesized video and reduce video coding. In order to solve this problem, in this article, the authors propose a spatiotemporal disparity refinement on local stereo matching based on the segmentation strategy. Based on segmentation information, matching point searching, and color similarity, adaptive disparity values to recover the disparity errors in disparity sequences can be obtained. The flickering errors are also effectively removed, and the boundaries of objects are well preserved. The procedures of the proposed approach consist of a segmentation process, matching point searching, and refinement in the temporal and spatial domains. Experimental results verify that the proposed approach can yield a high quantitative evaluation and a high-quality disparity map compared with other methods.


2014 ◽  
Vol 2014 ◽  
pp. 1-9
Author(s):  
Cheng-Tao Zhu ◽  
Yau-Zen Chang ◽  
Huai-Ming Wang ◽  
Kai He ◽  
Shih-Tseng Lee ◽  
...  

Developing matching algorithms from stereo image pairs to obtain correct disparity maps for 3D reconstruction has been the focus of intensive research. A constant computational complexity algorithm to calculate dissimilarity aggregation in assessing disparity based on separable successive weighted summation (SWS) among horizontal and vertical directions was proposed but still not satisfactory. This paper presents a novel method which enables decoupled dissimilarity measure in the aggregation, further improving the accuracy and robustness of stereo correspondence. The aggregated cost is also used to refine disparities based on a local curve-fitting procedure. According to our experimental results on Middlebury benchmark evaluation, the proposed approach has comparable performance when compared with the selected state-of-the-art algorithms and has the lowest mismatch rate. Besides, the refinement procedure is shown to be capable of preserving object boundaries and depth discontinuities while smoothing out disparity maps.


2020 ◽  
Vol 2020 ◽  
pp. 1-12 ◽  
Author(s):  
Kun Zhou ◽  
Xiangxi Meng ◽  
Bo Cheng

Stereo vision is a flourishing field, attracting the attention of many researchers. Recently, leveraging on the development of deep learning, stereo matching algorithms have achieved remarkable performance far exceeding traditional approaches. This review presents an overview of different stereo matching algorithms based on deep learning. For convenience, we classified the algorithms into three categories: (1) non-end-to-end learning algorithms, (2) end-to-end learning algorithms, and (3) unsupervised learning algorithms. We have provided a comprehensive coverage of the remarkable approaches in each category and summarized the strengths, weaknesses, and major challenges, respectively. The speed, accuracy, and time consumption were adopted to compare the different algorithms.


Sign in / Sign up

Export Citation Format

Share Document