Object-Centric Stereo Ranging for Autonomous Driving: From Dense Disparity to Census-Based Template Matching

2026-04-09

AI Summary

This work addresses the high computational cost, sensitivity to radiometric variations, and limited real-time performance of conventional dense stereo matching under long-range conditions and challenging illumination. To overcome these limitations, the authors propose a unified ranging framework that integrates object-level Census template matching with monocular geometric priors. Within detected bounding boxes, GPU-accelerated sparse matching is performed, enhanced by a suite of strategies including near-far disparity partitioning, forward-backward consistency checks, occlusion-aware sampling, and multi-block aggregation. An online extrinsic drift compensation module, based on radar voting and object association, further improves robustness. The system employs an asynchronous pipeline that seamlessly integrates Census transform, semi-global matching, and automatic image rectification, enabling robust and real-time long-range vehicle depth estimation in adverse conditions such as nighttime and rain, thereby significantly enhancing autonomous driving perception capabilities.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Accurate depth estimation is critical for autonomous driving perception systems, particularly for long range vehicle detection on highways. Traditional dense stereo matching methods such as Block Matching (BM) and Semi Global Matching (SGM) produce per pixel disparity maps but suffer from high computational cost, sensitivity to radiometric differences between stereo cameras, and poor accuracy at long range where disparity values are small. In this report, we present a comprehensive stereo ranging system that integrates three complementary depth estimation approaches: dense BM/SGM disparity, object centric Census based template matching, and monocular geometric priors, within a unified detection ranging tracking pipeline. Our key contribution is a novel object centric Census based template matching algorithm that performs GPU accelerated sparse stereo matching directly within detected bounding boxes, employing a far close divide and conquer strategy, forward backward verification, occlusion aware sampling, and robust multi block aggregation. We further describe an online calibration refinement framework that combines auto rectification offset search, radar stereo voting based disparity correction, and object level radar stereo association for continuous extrinsic drift compensation. The complete system achieves real time performance through asynchronous GPU pipeline design and delivers robust ranging across diverse driving conditions including nighttime, rain, and varying illumination.

Problem

• depth estimation
• stereo matching
• autonomous driving
• long-range detection
• disparity

Innovation

• object-centric stereo
• Census-based template matching
• online calibration refinement
• sparse stereo matching
• GPU-accelerated ranging