Fact-checked by Grok 2 weeks ago

Semi-global matching

Semi-global matching (SGM) is a algorithm designed for dense stereo matching, which estimates disparity maps from pairs of rectified stereo images to enable via . Introduced by Heiko Hirschmüller in 2005, it combines elements of local and global optimization techniques to achieve high accuracy while maintaining computational efficiency. The core of SGM involves computing a pixel-wise matching cost, typically using to compensate for radiometric differences between images, followed by aggregation of these costs along multiple one-dimensional paths (often in eight or sixteen directions) to enforce smoothness constraints and minimize discontinuities. This path-based optimization approximates a global energy minimization, outperforming purely local methods in handling slanted surfaces and object boundaries, while avoiding the high complexity of full global approaches like graph cuts or . Additional features include built-in handling, sub-pixel refinement for precise disparity estimation, and post-processing steps such as outlier filtering and gap , resulting in sub-pixel accuracy competitive with top-performing algorithms on benchmarks like the Middlebury Stereo Evaluation. With a linear time complexity of O(W × H × D)—where W and H are dimensions and D is the disparity —SGM processes typical images in 1-2 seconds and scales to large formats (up to billions of pixels) using techniques. It has been optimized for various platforms, including CPUs, GPUs (achieving 4.5 frames per second on 640×480 images), and FPGAs (up to 27 Hz with low power consumption), making it suitable for applications. Due to its robustness against illumination changes, low-texture regions, and parameter variations, SGM finds extensive use in fields such as , autonomous navigation, driver assistance systems, , and aerial matching.

Introduction

Overview

Semi-global matching (SGM) is a widely used algorithm in for estimating dense disparity maps from pairs of rectified stereo images, enabling accurate of scenes. It represents a approach that approximates the global optimization of disparity estimation—typically an NP-hard problem—by aggregating pixel-wise matching costs along multiple one-dimensional paths across the . This path-based aggregation enforces constraints in a semi-global manner, drawing on the interdependence of neighboring pixels modeled through Markov random fields. The innovation of SGM lies in its ability to balance the superior accuracy of global methods, which integrate information across the entire image to handle discontinuities and fine structures, with the efficiency of methods that operate on small windows. By optimizing costs directionally from various paths (such as , vertical, and diagonal), SGM achieves a polynomial-time that avoids the computational expense of full global minimization while producing high-quality results. Taking a rectified image pair as input—where corresponding are aligned along scanlines—SGM outputs a dense disparity assigning disparity values to each in the reference image, facilitating depth computation and applications like and autonomous . Key advantages include robust performance in textureless regions, where local methods often fail due to ambiguous matches, and improved handling of occlusions and radiometric variations, all while maintaining linear suitable for real-time processing on standard hardware.

Historical Development

Semi-global matching (SGM) was introduced by Heiko Hirschmüller in 2005 at the IEEE and (CVPR) conference in a paper titled "Accurate and Efficient Stereo Processing by Semi-Global Matching and ." Developed at the (DLR), the algorithm addressed key challenges in dense stereo matching, such as handling radiometric differences between images and achieving high accuracy at depth discontinuities while maintaining computational efficiency. The initial motivation stemmed from the need for robust stereo processing in planetary imaging applications, including the mapping of Mars surfaces using DLR's High Resolution Stereo Camera (HRSC), where traditional methods struggled with large-scale, low-texture scenes and varying illumination. In 2008, Hirschmüller published an extended journal version in IEEE Transactions on Pattern Analysis and Machine Intelligence, refining the method with pathwise cost aggregation and demonstrating its effectiveness on aerial and space imaging datasets, such as HRSC Mars images and urban aerial surveys. This publication solidified SGM's role in , enabling precise reconstructions for planetary exploration and . By this time, early implementations began appearing in research for , highlighting SGM's balance of and real-time feasibility. Adoption accelerated in the late 2000s, with re-implementations in open-source libraries like OpenCV's StereoSGBM module, officially integrated in version 2.1 in , facilitating its use in pipelines. This led to widespread application in and autonomous systems by the early 2010s, including stereo vision for planetary rovers at . Milestone hardware integrations emerged around the same period, such as GPU accelerations in 2008 for real-time processing and FPGA implementations in 2009 for low-power embedded systems, paving the way for deployment in resource-constrained environments like space missions.

Background

Stereo Matching Fundamentals

Stereo vision is a technique that estimates the three-dimensional structure of a scene by analyzing two or more images captured from slightly different viewpoints, typically using parallel cameras to exploit for depth recovery. describes the projective relationship between corresponding points in the images, where a point in one image projects onto an epipolar line in the other, constraining the search for matches and reducing the correspondence problem from 2D to 1D. Disparity refers to the horizontal pixel shift between corresponding points in the left and right images of a rectified pair, serving as an inverse measure of depth. The depth Z at a point can be computed from the disparity d using the formula Z = \frac{f \cdot b}{d}, where f is the of the camera and b is the distance between the two cameras. This relationship assumes a and calibrated, rectified images, enabling direct conversion from disparity maps to depth information for . Image rectification is a preprocessing step that transforms the stereo images so that epipolar lines become horizontal and aligned across both views, simplifying the matching process to a one-dimensional search along corresponding rows. This transformation is achieved by applying homographies derived from the camera calibration parameters, ensuring that corresponding points share the same vertical coordinate. Stereo matching algorithms can be categorized as sparse or dense, with dense methods computing disparity values for every in the image to produce complete disparity maps suitable for applications like 3D surface reconstruction and view interpolation. In contrast, sparse matching focuses on distinctive features such as edges or corners, often requiring post-processing to interpolate depths across the entire image, but dense approaches provide more comprehensive coverage for dense pixel-wise depth estimation.

Limitations of Prior Methods

Prior to the development of semi-global matching, stereo matching algorithms were broadly categorized into local and global approaches, each exhibiting significant trade-offs in accuracy, efficiency, and robustness. Local methods, such as winner-takes-all schemes with window-based similarity measures (e.g., sum of squared differences or normalized ), prioritize computational speed by aggregating matching costs within fixed or adaptive windows around each . These techniques achieve performance, often processing images in under 2 seconds on contemporary , making them suitable for applications requiring low latency. However, they suffer from high to and illumination variations, as pixelwise cost computations are inherently ambiguous and prone to erroneous matches in low-texture regions. A primary limitation of local methods is the production of streaking artifacts in textureless areas, where the assumption of constant disparity within the aggregation window leads to inconsistent estimates across scanlines, resulting in visible linear patterns in the disparity map. Additionally, these methods fail at depth discontinuities, causing over-smoothing and "foreground fattening" effects that blur object boundaries and fine structures, as the window-based aggregation propagates neighboring disparities inappropriately. Occlusions exacerbate these issues, often inducing front-parallel biases where hidden regions are incorrectly assigned disparities from visible surfaces, further degrading accuracy in uniform or occluded scenes. In contrast, global methods, including graph cuts, , and dynamic programming, aim for higher accuracy by minimizing a 2D function that enforces smoothness constraints across the entire , yielding disparity maps that are more consistent and less prone to artifacts in textureless regions. These approaches excel at handling depth discontinuities and occlusions through explicit modeling in the energy terms, producing smoother results with reduced over-smoothing compared to local techniques. However, their computational intensity is a major drawback; for instance, graph cuts and often exhibit O(N^2) or higher complexity relative to image size N, leading to runtimes of 20 seconds to several minutes per image, alongside substantial memory demands for storing graph representations or message-passing states. Global methods also remain sensitive to radiometric changes, such as varying illumination between pairs, requiring careful design to avoid biased matches, and they can introduce in dynamic programming variants due to independent 1D optimizations per scanline. While capable of sub-pixel accuracy in controlled benchmarks, their high resource requirements render them unsuitable for applications or large-scale processing, such as in or autonomous driving. These trade-offs—efficiency but inaccuracy in local methods versus accuracy but slowness in global methods—highlighted the need for approaches that approximate global optimality with manageable computational costs.

Core Algorithm

Matching Cost Computation

In semi-global matching (SGM), the matching cost computation forms the foundation of the disparity estimation process by quantifying the dissimilarity between corresponding pixels in a rectified stereo image pair. This pixelwise cost is essential for handling radiometric differences, such as variations in illumination or sensor characteristics between the base image I_b and the match image I_m. The original SGM algorithm employs (MI) as the primary matching cost metric, which measures the statistical dependence between intensity distributions without assuming a linear relationship between pixel values. The MI-based cost for a pixel p in the base image and disparity d is defined as C_{MI}(p, d) = -MI(I_b, f_D(I_m))(I_b^p, I_m^q), where q = e_{bm}(p, d) denotes the epipolar line in the match image, f_D is a warping based on disparity d, and MI is computed from joint and marginal entropies: MI(I_1, I_2) = H(I_1) + H(I_2) - H(I_1, I_2). Entropies are estimated using discretized 256-bin histograms of intensities, smoothed with a Gaussian kernel (e.g., 7×7) to reduce sensitivity. To address the computational expense of full MI calculation, a hierarchical (HMI) is used: it starts with a coarse disparity estimate from downsampled images and refines it iteratively, adding approximately 14% to the runtime while maintaining accuracy. For robustness in radiometrically varying scenes, the MI cost avoids simplistic metrics like absolute or squared differences, which fail under non-linear transformations. An alternative in the original implementation is the Birchfield-Tomasi () cost, which computes a sampling-insensitive measure of difference by linearly interpolating pixel values along epipolar lines, formulated as the minimum of forward and backward warped differences: C_{BT}(p, d) = \min(|I_b^p - I_m^{q_f}|, |I_b^p - I_m^{q_b}|, |I_b^p - I_m^q|), where q_f and q_b account for subpixel shifts. This cost is particularly effective for scenes with occlusions or textureless regions when combined with a small support window (e.g., 5×5 ). Preprocessing assumes rectified images with known , though extensions allow for unrectified pairs by incorporating geometric transformations. In practice, the matching cost is aggregated over multiple disparity hypotheses (typically 64–128 levels) and a local neighborhood to form an initial cost volume, which is then refined through path-based aggregation in subsequent SGM steps. This design ensures subpixel accuracy (e.g., 0.25- ) and low error rates on datasets like the Middlebury stereo evaluation, where MI-based costs yield bad-pixel errors below 10% in textured regions.

Path-Based Cost Aggregation

In semi-global matching, path-based cost aggregation approximates global smoothness constraints by propagating matching costs along multiple one-dimensional paths across the image, rather than performing exhaustive two-dimensional optimization. This step transforms the initial pixelwise matching cost volume C(p, d)—computed for each pixel p and disparity d—into a smoothed aggregate that encourages smooth disparity fields while preserving edges. By limiting aggregation to linear paths, the achieves computational efficiency while closely approximating the minimum-energy solution of a global stereo model. Paths are selected to cover the two-dimensional support region of each pixel without requiring a full graph-based computation. Typically, at least eight directions are used, such as horizontal, vertical, and diagonal paths originating from the image borders, with sixteen directions recommended for better approximation of global optimality. These paths are straight lines in the reference image but may appear non-straight in the matching image due to varying disparities along the way. For each path direction r, dynamic programming is applied sequentially from the path's starting border pixel toward the target pixel p, computing the minimum cost L_r(p, d) that reaches p at disparity d. This recursive formulation penalizes abrupt disparity changes to enforce smoothness, allowing small variations for slanted surfaces while imposing higher costs on large jumps indicative of depth discontinuities. The core dynamic programming update for the cost along path r is given by L_r(p, d) = C(p, d) + \min \left( L_r(p - r, d),\ L_r(p - r, d - 1) + P_1,\ L_r(p - r, d + 1) + P_1,\ \min_i L_r(p - r, i) + P_2 \right) - \min_k L_r(p - r, k), where p - r denotes the predecessor along the path, P_1 is a small constant penalty for one-pixel disparity changes, and P_2 (with P_2 \geq P_1) is a larger penalty for changes exceeding one pixel, often adapted to gradients to better preserve edges. The subtraction of \min_k L_r(p - r, k) normalizes the costs to prevent unbounded growth, ensuring L_r(p, d) \leq C_{\max} + P_2 where C_{\max} is the maximum initial matching cost. This formulation balances fidelity to evidence with , as the minimum over predecessor states favors smooth propagation unless a discontinuity justifies a higher penalty. The final aggregated cost for pixel p at disparity d is obtained by summing the minimum costs over all path directions: S(p, d) = \sum_r L_r(p, d). This approximates the energy minimization of a model for , where the -based approach reduces the from quadratic in image dimensions to linear, specifically O(W \cdot H \cdot D \cdot N) with image width W, height H, maximum disparity D, and N paths (typically 8–16), making it suitable for applications. The resulting S(p, d) provides a robust energy landscape for subsequent disparity selection, effectively mitigating the streaking artifacts common in purely local methods.

Disparity Map Selection

In the final step of semi-global matching, the disparity map is generated through winner-takes-all (WTA) , where for each p, the optimal disparity d^* is selected as the value that minimizes the aggregated matching : d^* = \arg\min_d S(p, d), with S(p, d) representing the total cost aggregated across multiple paths from the preceding aggregation stage. This produces an initial disparity image D_b that assigns an integer disparity to every based solely on the lowest , enabling efficient without iterative optimization. Post-processing refines this initial to remove artifacts and handle inconsistencies. A 3×3 is applied to D_b (and a corresponding right-to-left D_m) to eliminate speckle and outliers, preserving edges while smoothing isolated erroneous disparities. Additionally, a left-right check detects occlusions and mismatches by comparing corresponding disparities: for a p in the left with disparity D_b, the corresponding point q in the right is evaluated, and if |D_b - D_m| > 1, the disparity is invalidated to mark occluded or unreliable regions. To achieve sub-pixel accuracy beyond integer disparities, quadratic interpolation is performed around the WTA minimum. This fits a parabola to the aggregated costs at the minimum disparity and its two neighbors, estimating the precise sub-pixel offset that minimizes the and thereby enhancing the overall of the . The resulting output is a dense disparity , where invalid regions from the consistency check may be interpolated for completeness, providing a comprehensive representation of depth that can be triangulated with camera parameters to yield a 3D for applications such as .

Variants and Extensions

Memory-Efficient Variant

The standard Semi-global matching (SGM) algorithm faces significant memory challenges due to the cost volume, which requires O(W × H × D) storage, where W and H are the width and , and D is the disparity range; for high-resolution images such as 1 megapixel with D=128, this can exceed 1 , making it prohibitive for embedded systems or large-scale processing. To address this, early adaptations introduced strategies, processing the image in overlapping strips or tiles to limit usage while maintaining boundary consistency through weighted merging of results from adjacent tiles. Additionally, using 16-bit integers instead of floating-point representations for cost storage further reduces the footprint, scaling costs to fit within 11 bits for the initial matching costs. A more advanced memory-efficient variant, known as efficient SGM (eSGM), pipelines the path aggregation process to avoid storing the full cost volume by computing forward and backward passes in-place along sequential directions, such as top-down and bottom-up stripes, thereby reducing the temporary requirement to O(W × D). This approach reuses by overwriting minima from previous aggregation passes and limits the number of paths—typically to four or eight instead of the full set in standard SGM—to prioritize and vertical directions while aggregating costs sequentially (e.g., paths first, followed by vertical). Such pipelining enables processing of high-resolution images on resource-constrained , including GPUs and FPGAs, achieving performance; for instance, it processes 640×480 images with D=64 in 0.06 seconds on FPGAs. While eSGM incurs a slight accuracy , with error rates on datasets like Middlebury differing by less than 0.1% compared to standard SGM (e.g., 7.16% vs. 7.17% bad error), it increases computational time by approximately 50% due to the multi-pass nature but remains suitable for applications requiring low memory, such as automotive vision. This variant builds on the path-based aggregation of standard SGM by adapting it for sequential reuse rather than simultaneous computation across all paths.

Advanced Improvements

Since its inception, Semi-global matching (SGM) has undergone significant enhancements to address limitations in handling complex scenes, such as untextured regions and varying lighting conditions, particularly in post-2010 developments focused on accuracy and adaptability. One notable advancement is the Semi-Global Matching with Priors (SGMP) algorithm introduced in , which incorporates surface orientation priors to better manage slanted untextured surfaces that traditional SGM struggles with due to over-smoothed disparities. By adding geometric constraints directly to the energy function, SGMP enforces piecewise-planar assumptions on surface normals estimated from initial disparity cues, resulting in reduced errors on benchmarks like the Middlebury dataset, where it achieves up to 20% improvement in bad rates for slanted areas compared to vanilla SGM. Building on these geometric refinements, the Improved Semi-Global Matching (I-SGM) variant, proposed in , tailors SGM for challenging environments like lunar rover navigation. I-SGM enhances edge preservation by introducing adaptive penalties that dynamically adjust based on local magnitudes and illumination variations, mitigating artifacts in low-contrast, dimly lit terrains with uneven shadows. This adaptation proves particularly effective for obstacle detection, yielding disparity maps with 15-25% fewer outliers in simulated lunar datasets under complex lighting, thereby improving rover path planning reliability. Adaptations for neuromorphic sensors represent another key evolution, with event-based SGM emerging around 2018 to leverage dynamic vision sensors (DVS) like event cameras. Unlike frame-based SGM, event-based versions process asynchronous change events rather than full images, enabling sub-millisecond for in high-speed or dynamic scenes. By accumulating events into contrast-maximized representations and applying path aggregation on these sparse inputs, the method preserves depth edges in motion-blurred environments, achieving performance at over Hz on synthetic and real DVS datasets while reducing motion artifacts by up to 30% compared to traditional . Hardware accelerations have further propelled SGM's practicality for applications, exemplified by FPGA implementations in 2023 that employ comparator structures to expedite cost aggregation. These designs pipeline multiple disparity hypotheses across systolic arrays, minimizing accesses and enabling of HD-resolution (1280×720) stereo pairs at up to 60 frames per second on FPGAs such as Stratix V, with power consumption under 2 on Zynq UltraScale+. Such optimizations maintain SGM's accuracy—nearly identical to software versions—while enabling suitability for systems in and automotive vision. By 2025, recent trends in SGM enhancements increasingly integrate for robust cost initialization, with hybrid SGM-CNN models combining neural feature extractors to initialize matching costs before classical aggregation. These hybrids, such as those using CNNs for radiometric-invariant descriptors, enhance disparity accuracy in scenarios with illumination discrepancies or textureless regions, reporting 10-15% gains in endpoint error on KITTI benchmarks over pure SGM. This fusion leverages CNNs' learned invariance while retaining SGM's global consistency, fostering deployment in diverse real-world conditions like autonomous driving.

Applications and Evaluation

Key Applications

Semi-global matching (SGM) has found widespread adoption in and autonomous vehicles, where it enables robust depth sensing for obstacle avoidance and 3D environmental mapping. In planetary exploration, the () has integrated SGM into rover systems since 2005, utilizing stereo cameras to generate dense disparity maps for safe navigation on uneven terrains, as demonstrated in the IDEFIX rover's autonomous navigation experiments. Similarly, in advanced driver-assistance systems (ADAS), SGM supports stereo vision for , aiding in features like and lane departure warnings by providing accurate depth information from vehicle-mounted cameras. In aerial and satellite imagery processing, SGM excels at deriving digital surface models (DSMs) from stereo pairs, facilitating applications such as in orthoimages and urban modeling. For instance, it has been employed to monitor environmental changes, like forest area alterations, by computing precise disparity maps from satellite stereo data, enabling the identification of temporal differences in elevation. In urban settings, drone-captured stereo imagery processed with SGM supports the creation of detailed city models, capturing building facades and street layouts for applications in and infrastructure . Medical imaging benefits from SGM in 3D reconstruction tasks, particularly with stereo endoscopes during minimally invasive surgeries. The algorithm processes binocular endoscopic images to estimate depth maps, providing surgeons with enhanced spatial awareness of internal anatomies, which improves precision in procedures like by visualizing tissue surfaces in three dimensions. Consumer devices increasingly incorporate SGM variants for (AR) and (VR) functionalities. In smartphones, semi-global block matching—a derivative of SGM—utilizes dual-camera setups to compute depth for features like portrait mode bokeh effects and AR object placement, enabling seamless integration of virtual elements into real-world scenes captured by mobile lenses. VR headsets leverage similar techniques for environment mapping, generating disparity maps from stereo views to support immersive pass-through video and spatial tracking. Emerging applications of SGM extend to and high-speed . Improved variants like I-SGM enhance obstacle detection on lunar and Martian surfaces under challenging illumination, as tested for in complex environments. Additionally, event-based adaptations of SGM, which process asynchronous visual data from neuromorphic sensors, enable low-latency depth estimation for high-speed robotic tasks, such as agile manipulation in dynamic settings, with ongoing developments enabling near-real-time performance on specialized hardware as of 2023.

Performance Comparisons

Semi-global matching (SGM) demonstrates strong performance in stereo matching benchmarks, particularly in balancing accuracy and computational efficiency. On earlier Middlebury stereo datasets (pre-2014), standard SGM implementations achieve bad pixel errors (pixels with disparity error >1 pixel) in the range of 5-7%, with variants like improved SGM reaching as low as 4.1%; on the more challenging dataset, standard SGM errors are around 25-30%, though variants achieve 4-7%. This outperforms traditional local methods that often exceed 20% error due to poor handling of occlusions and textureless regions. For sub-pixel , SGM's endpoint error averages around 0.5-1.0 pixels on Middlebury , providing reliable dense disparity maps for applications requiring fine detail. In comparisons to local methods like block matching, SGM improves accuracy in textureless areas by approximately 50%, reducing erroneous matches from over 30% to under 15% in such regions. Versus global methods such as graph cuts, SGM is roughly 10 times faster while maintaining comparable accuracy, as graph cuts require iterative optimization over the full energy function. However, SGM is slower than basic block matching, which processes images in milliseconds but at the cost of higher error rates (e.g., 15-25% bad pixels on Middlebury). Deep learning-based methods like PSMNet achieve superior accuracy, with bad pixel errors as low as ~3.1% (Out-Noc) on the KITTI benchmark compared to typical SGM variants' 9-15%, but they are 100 times slower on CPU (minutes per image versus SGM's seconds) and require GPU acceleration for use. Standard SGM runtime is 0.1-1 second per image on a modern CPU for VGA-resolution pairs, enabling near- processing, while memory-efficient variants run in under 30 ms on embedded hardware like FPGAs. SGM proves robust to radiometric differences and but remains sensitive to penalty (P1 for small disparity changes, P2 for large jumps), where suboptimal values can increase errors by 20-30%. Evaluated on standard datasets like Middlebury (for controlled indoor scenes) and KITTI (for outdoor driving), SGM consistently ranks high in the efficiency-accuracy trade-off, as per Middlebury 2014 and KITTI 2012/2020 benchmarks, where optimized variants excel in non-occluded regions with ~5-10% error. Despite these strengths, SGM struggles with thin structures and repetitive patterns, leading to over 15% error in such cases; recent hybrids combining CNN-based cost computation with SGM aggregation mitigate this, reducing errors on thin objects by up to 40% in 2020s benchmarks.
MethodBad Pixel Error (Middlebury, %)Bad Pixel Error (KITTI D1-all, %)Runtime (CPU, VGA image)
Local Block Matching20+15-25<0.1 s
SGM (Standard, pre-2014)5-7~220.1-1 s
SGM (Optimized Variants)4-7 (2014)9-150.1-1 s
Graph Cuts4-67-101-10 s
PSMNet (Deep Learning)3-4~2.8>100 s (no GPU)

References

  1. [1]
    [PDF] Stereo Processing by Semi-Global Matching and Mutual Information
    This paper describes the Semi-Global Matching (SGM) stereo method. It uses a pixelwise, Mutual Information based matching cost for compensating radiometric ...Missing: original | Show results with:original
  2. [2]
    [PDF] Accurate and Efficient Stereo Processing by Semi-Global Matching ...
    This paper considers the objectives of accurate stereo matching, especially at object boundaries, robustness against recording or illumination changes and ...
  3. [3]
    [PDF] Semi-Global Matching – Motivation, Developments and Applications
    This paper motivates and explains the method, shows current developments as well as examples from various applications. INTRODUCTION. Stereo matching is used ...
  4. [4]
    [PDF] An Introduction to the SGM Algorithm for Dense Matching in ...
    the following publication falls in that category: “Stereo Processing by Semiglobal Matching and Mutual. Information” by Heiko Hirschmüller, IEEE-PAMI, 2008.
  5. [5]
    Stereo Disparity Using Semi-Global Block Matching - MathWorks
    This example shows how to compute disparity between left and right stereo camera images using the Semi-Global Block Matching algorithm.
  6. [6]
    Accurate and efficient stereo processing by semi-global matching ...
    Accurate and efficient stereo processing by semi-global matching and mutual information ; Article #: ; Date of Conference: 20-25 June 2005 ; Date Added to IEEE ...
  7. [7]
    LRU
    A stereo camera and multi-award winning Semi-Global-Matching stereo vision ... In 2015, the LRU took part in the SpaceBot Camp of the DLR Space Administration.
  8. [8]
    [PDF] A Taxonomy and Evaluation of Dense Two-Frame Stereo ...
    Scharstein and R. Szeliski. A taxonomy and evaluation of dense two-frame stereo correspondence algorithms. Tech- nical Report MSR-TR-2001-81 ...
  9. [9]
    [PDF] Epipolar Geometry and the Fundamental Matrix
    The epipolar geometry is the intrinsic projective geometry between two views. It is independent of scene structure, and only depends on the cameras' internal ...
  10. [10]
    [PDF] A Compact Algorithm for Rectification of Stereo Pairs - Andrea Fusiello
    Any pair of images can be transformed so that epipolar lines are parallel and horizontal in each image. This procedure is called rectification. 3 Rectification ...
  11. [11]
    Stereo Processing by Semiglobal Matching and Mutual Information
    Feb 29, 2008 · This paper describes the semiglobal matching (SGM) stereo method. It uses a pixelwise, mutual information (Ml)-based matching cost for compensating radiometric ...
  12. [12]
  13. [13]
    [PDF] MEMORY EFFICIENT SEMI-GLOBAL MATCHING
    The Semi-Global Matching (SGM) method (Hirschmüller, 2008) minimizes a global cost function only in one dimension like Dy- namic Programming, but the direction ...<|control11|><|separator|>
  14. [14]
    Semi-Global Stereo Matching with Surface Orientation Priors - arXiv
    Semi-Global Matching (SGM) is a widely-used efficient stereo matching technique. It works well for textured scenes, but fails on untextured ...Missing: SGMP | Show results with:SGMP
  15. [15]
    Research on the Improvement of Semi-Global Matching Algorithm ...
    Aug 3, 2023 · ... I-SGM algorithm is applicable in dim conditions on the lunar surface and can better help a lunar rover to detect obstacles during its travel.
  16. [16]
    Event-based stereo matching using semiglobal matching
    Jan 17, 2018 · In this article, we present a fully event-based stereo matching algorithm for reliable 3-D depth estimation using semiglobal matching (SGM).Missing: SGMP | Show results with:SGMP
  17. [17]
    A Hardware Accelerator for the Semi-Global Matching Stereo ...
    In this work, a novel hardware implementation of the semi-global matching algorithm is presented. A hardware structure of parallel comparators is proposed for ...
  18. [18]
    Review of Stereo Matching Algorithms Based on Deep Learning
    Mar 23, 2020 · The SGM-method suffers from obvious bad pixels while the GC-Net provides a much smoother and more consistent depth map. A detailed review of the ...
  19. [19]
    [PDF] Strong Stereo Features for Self-Supervised Practical Stereo Matching
    Our proposed approach is a novel self-supervised learning-based SGM method. This is a promising hybrid[47] pathway to more explainable deep stereo and industry ...
  20. [20]
    [PDF] The DLR Autonomous Navigation Experiment with the IDEFIX Rover
    methods like in NASA's Mars Exploration Rovers [20], we use the Semi-global matching algorithm (SGM) [21] with rectified stereo images. SGM was developed at ...
  21. [21]
    [PDF] Change Detection Analysis of Forest Areas Using Satellite Stereo Data
    The Semi Global Matching (SGM) algorithm by [Hi08] is used to per- form dense and reliable stereo matching. SGM avoids using matching windows, and is thus able ...
  22. [22]
    Depth estimation and 3D reconstruction from UAV-borne imagery
    Initially intended for two-view stereo depth estimation, the so-called Semi-Global Matching (SGM) algorithm was proposed by Hirschmueller (2005, 2008). Due to ...
  23. [23]
    Dense Depth Estimation from Stereo Endoscopy Videos Using ... - NIH
    Estimating depth in stereo endoscopy can be achieved via densely matching pixels from a pair of binocular images. ... semi-global matchingand mutual information.
  24. [24]
    [PDF] Real-Time Depth Estimation Using On-Device Dual Cameras
    f for the dual cameras. Semi-Global Block Matching (SGBM) SGBM is a commonly used computer vision algorithm in binocular camera systems for depth estimation.
  25. [25]
    (PDF) Event-based stereo matching using semiglobal matching
    We propose a fully event-based stereo three-dimensional depth estimation algorithm inspired by semiglobal matching. Our algorithm considers the smoothness ...
  26. [26]
    [PDF] Real-time Stereo Vision: Optimizing Semi-Global Matching - ini.rub.de
    Hirschmüller, “Accurate and efficient stereo processing by semi- global matching and mutual information,” in Proceedings of the IEEE. Conference on Computer ...
  27. [27]
    Middlebury Stereo Evaluation - Version 3
    Learning to fuse proposals from multiple scanline optimizations in semi-global matching. ECCV 2018. Semi-Global Matching (SGM) uses an aggregation scheme to ...
  28. [28]
    [PDF] High-Resolution Stereo Datasets with Subpixel-Accurate Ground Truth
    Stereo algorithm performance on our new datasets. Top: Bad pixel percentages. (% error > 1.0) for Census, ELAS, and SGM, sorted by SGM error, for imperfect.
  29. [29]
    18. The stereo algorithms in ASP in detail
    Semi-Global Matching and More Global Matching algorithms¶. ASP implements the popular Semi-Global Matching (SGM) algorithm introduced in [Hirschmuller08], and ...<|control11|><|separator|>
  30. [30]
    [PDF] arXiv:1409.4326v2 [cs.CV] 20 Oct 2015
    Oct 20, 2015 · We describe how a convolutional neural network can be used to compute the stereo matching cost. • We achieve an error rate of 2.61% on the KITTI.
  31. [31]
    [PDF] SGM: Real-time Raster-Respecting Semi-Global Matching for Power ...
    On KITTI, we compare our approach to the only two published approaches from the benchmark that are able to achieve state-of-the-art performance in real time [15] ...
  32. [32]
    [PDF] arXiv:1904.06587v1 [cs.CV] 13 Apr 2019
    Apr 13, 2019 · First, SGM has many user-defined parameters (P1,P2), which are not straightforward to tune. All of these param- eters become unstable ...
  33. [33]
    Stereo Evaluation 2012 - The KITTI Vision Benchmark Suite
    Our evaluation server computes the average number of bad pixels for all non-occluded or occluded (=all groundtruth) pixels. We require that all methods use the ...Missing: speed | Show results with:speed
  34. [34]
    [PDF] Stereo Matching by Training a Convolutional Neural Network to ...
    Heiko Hirschmüller. Stereo processing by semiglobal matching and mutual information. IEEE Transactions on Pattern Analysis and Machine Intelligence, 30(2): ...
  35. [35]
    [PDF] A Comparative Study on Deep-Learning Methods for Dense Image ...
    This paper evaluates four DL stereo matching methods: MC-CNN, GCNet, PSMNet, and LEAStereo, against Census-SGM, using multi-date satellite stereo pairs.