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

@inproceedings{Pinggera2014KnowYL,
  title={Know Your Limits: Accuracy of Long Range Stereoscopic Object Measurements in Practice},
  author={Peter Pinggera and David Pfeiffer and Uwe Franke and Rudolf Mester},
  booktitle={ECCV},
  year={2014}
}
Modern applications of stereo vision, such as advanced driver assistance systems and autonomous vehicles, require highest precision when determining the location and velocity of potential obstacles. Subpixel disparity accuracy in selected image regions is therefore essential. 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 treat local… 
A unified method for improving long-range accuracy of stereo and monocular depth estimation algorithms
TLDR
This work introduces a taxonomy to categorize all types of camera-based depth perception methods with respect to their long-range capabilities, and presents a correction method that works for both stereo and monocular depth perception algorithms that output a depth in discrete setting.
Improving stereo reconstruction by sub-pixel correction using histogram matching
TLDR
A novel sub-pixel disparity correction technique based on applying histogram matching through the use of generated Look-up Tables (LUTs) that is flexible, fast and produces more accurate results than previous solutions in the discrete domain.
New sub-pixel interpolation functions for accurate real-time stereo-matching algorithms
TLDR
A tool that provides qualitative measurements of the fitted function by the use of generated ground truth synthetic images, with sub-pixel precision is implemented and the best interpolation functions for the SGM algorithms have been found and they outperform the state of the art functions, thus the proposed tool is validated.
Stereovision Bias Removal by Autocorrelation
  • Yang Cheng, L. Matthies
  • Environmental Science
    2015 IEEE Winter Conference on Applications of Computer Vision
  • 2015
TLDR
Stereovision Bias Removal by Autocorrelation (SBRA) is described, achieving 0.02 pixel RMS disparity error in synthetic stereo image data and a significant error reduction on real stereo images for which no ground truth is available.
Hierarchical Deep Stereo Matching on High-Resolution Images
TLDR
It is demonstrated that the performance-vs-speed tradeoff afforded by on-demand hierarchies may address sensing needs for time-critical applications such as autonomous driving.
Semantic segmentation-based stereo reconstruction with statistically improved long range accuracy
TLDR
This work proposes a new real-time stereo reconstruction method that uses a ConvNet for semantically segmenting the driving scene and improves the sub-pixel accuracy of the stereo matching by finding optimal interpolation functions for each particular segment class.
Fast robust monocular depth estimation for Obstacle Detection with fully convolutional networks
TLDR
This work proposes a novel appearance-based Object Detection system that is able to detect obstacles at very long range and at a very high speed (~ 300Hz), without making assumptions on the type of motion.
Volumetric Correspondence Networks for Stereo Matching and Optical Flow
TLDR
This thesis investigates efficient and effective ways of incorporating cost volume processing into deep neural networks for correspondence tasks and shows that for stereo matching on high-resolution images, 3D cost volumes can be efficiently filtered in a coarse-to-fine manner, and for optical flow estimation, ”true” 4D volumetric processing can be effectively utilized to improve model's accuracy as well as generalization ability.
Robust and accurate 3D motion estimation under adverse conditions
TLDR
This thesis considers the estimation of scene flow, the joint computation of motion and geometry from images of (calibrated) stereo cameras, acquired over at least two time steps and shows that scene flow methods can indeed outperform dedicated stereo and optical methods at their respective task.
Estimating the 2D Static Map Based on Moving Stereo Camera
TLDR
The aim of this study is to provide a solution based on the low-cost, light, and low-power stereo camera that focuses on the spatial information about the driving environment which is represented as a 3D point cloud.
...
1
2
3
4
...

References

SHOWING 1-10 OF 44 REFERENCES
Improving Stereo Sub-Pixel Accuracy for Long Range Stereo
  • S. Gehrig, Uwe Franke
  • Environmental Science, Computer Science
    2007 IEEE 11th International Conference on Computer Vision
  • 2007
TLDR
This work introduces a new disparity smoothing algorithm that preserves depth discontinuities and enforces smoothness on a sub-pixel level, and presents a novel stereo constraint (gravitational constraint) that assumes sorted disparity values in vertical direction and guides global algorithms to reduce false matches, especially in low-textured regions.
Highly Accurate Depth Estimation for Objects at Large Distances
TLDR
The proposed approach outperforms state-of-the-art stereo methods, achieving an average object disparity RMS error below 0.1 pixel, at typical object sizes of less than 15x15 pixels.
How Accurate Can Block Matches Be in Stereo Vision?
TLDR
The subpixel accuracy attainable for the disparity computed from a rectified stereo pair of images with small baseline is explored, and a formula for the variance of the disparity error caused by the noise is introduced and proved.
Pushing the limits of stereo using variational stereo estimation
TLDR
A novel stereo model is introduced that features a second-order regularization, which both allows sub-pixel accurate solutions and piecewise planar disparity maps and incorporates a robust fidelity term to account for adverse illumination conditions that frequently arise in real-world scenes.
Exploiting the Power of Stereo Confidences
TLDR
This paper makes full use of the stereo confidence cues by propagating all confidence values along with the measured disparities in a Bayesian manner and shows that using stereoconfidence cues allows both reducing the number of false object detections by a factor of six while keeping the detection rate at a near constant level.
Are we ready for autonomous driving? The KITTI vision benchmark suite
TLDR
The autonomous driving platform is used to develop novel challenging benchmarks for the tasks of stereo, optical flow, visual odometry/SLAM and 3D object detection, revealing that methods ranking high on established datasets such as Middlebury perform below average when being moved outside the laboratory to the real world.
Design of Interpolation Functions for Subpixel-Accuracy Stereo-Vision Systems
TLDR
This work proposes two methodologies for interpolation function generation based on data generated by the stereo algorithms that employs synthetic images of a known environment and applies function fitting to the resulted data.
A taxonomy and evaluation of dense two-frame stereo correspondence algorithms
TLDR
A taxonomy of dense, two-frame stereo methods designed to assess the different components and design decisions made in individual stereo algorithms is presented.
Improved sub-pixel stereo correspondences through symmetric refinement
TLDR
It is shown that the symmetric sub-pixel refinement strategy results in more accurate correspondences by avoiding bias while preserving detail, and is demonstrated to be effective in dense stereo correspondence algorithms.
Meaningful Matches in Stereovision
TLDR
A contrario block-matching method that cannot rule out false matches due to the presence of periodic objects in the images, but is successfully complemented by a parameterless self-similarity threshold.
...
1
2
3
4
5
...