Related Experiment Video
Updated: Jun 19, 2026
06:25
Stereoacuity Improvement using Random-Dot Video Games
Published on: January 14, 2020
Efficient stereoscopic ranging via stochastic sampling of match quality
Thayne Richard Coffman1, Alan Conrad Bovik
121st Century Technologies, Austin, TX 78759 USA. tcoffman@21technologies.com
Summary
This study introduces an efficient stereo matching method using stochastic sampling for accurate depth estimation. The approach improves video processing and outperforms existing methods in accuracy and resource efficiency.
Area of Science:
- Computer Vision
- Artificial Intelligence
Background:
- Dense stereo correspondence is crucial for 3D scene reconstruction.
- Existing methods often face challenges with accuracy, computational cost, and memory usage.
Purpose of the Study:
- To develop an efficient and accurate method for computing dense stereo correspondences.
- To improve depth estimation in stereo vision, particularly for video sequences.
Main Methods:
- Stochastic sampling of match quality values for non-integer disparities.
- Iterative refinement using stochastic cooperative search and perturbation.
- Seeding initial estimates in video using a novel Z-buffering algorithm application.
Main Results:
- The proposed method achieves higher accuracy than microcanonical annealing and cooperative approaches.
- It requires fewer match quality evaluations compared to benchmark methods.
- Demonstrates superior memory usage and scaling properties over exhaustive sampling alternatives.
Conclusions:
- The stochastic sampling approach offers an efficient and accurate solution for dense stereo correspondence.
- The method shows significant advantages for real-time video applications due to reduced computational demands.
