Scholarlink

<p>Accurate vehicle speed estimation is essential for forensic traffic accident reconstruction, yet conventional video-based methods typically require prior camera calibration or spatial road data. This study introduces a modular framework that estimates vehicle speed from fixed-camera footage even when parts of the vehicle trajectory are completely occluded, without any prior information about the camera or accident scene. The framework integrates SiamRPN-based single-vehicle tracking with Kalman filtering to handle occlusions and uses Mask R-CNN instance segmentation to extract wheel centers. It then applies a smoothed unscented Kalman filter to predict occluded wheel coordinates and computes vehicle speed from the geometric cross-ratio—a perspective-invariant distance ratio—combined with known wheelbase specifications. Validation across three settings demonstrated high accuracy: simulations on straight and curved roads yielded maximum errors of 0.30&#xa0;km/h and 1.70&#xa0;km/h, respectively; real vehicle tests achieved errors of 0.41&#xa0;km/h under constant-speed driving and 0.57&#xa0;km/h during acceleration; and actual accident case studies produced errors of 0.49–3.05&#xa0;km/h compared with field measurements. Unlike existing approaches that provide only point or interval estimates, the proposed framework generates continuous speed profiles throughout the accident sequence, enabling dynamic analysis of driver behavior such as acceleration, deceleration, and braking response. When vehicle specifications are available, this approach delivers legally defensible speed estimates from readily available video evidence, significantly improving the precision and practical utility of forensic accident reconstruction.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Robust video-based vehicle speed estimation for occluded scenes for forensic analysis of traffic accidents

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月2日

查看全文
  • Youngsoo Choi,
  • Seung-Hyun Kong

摘要

Accurate vehicle speed estimation is essential for forensic traffic accident reconstruction, yet conventional video-based methods typically require prior camera calibration or spatial road data. This study introduces a modular framework that estimates vehicle speed from fixed-camera footage even when parts of the vehicle trajectory are completely occluded, without any prior information about the camera or accident scene. The framework integrates SiamRPN-based single-vehicle tracking with Kalman filtering to handle occlusions and uses Mask R-CNN instance segmentation to extract wheel centers. It then applies a smoothed unscented Kalman filter to predict occluded wheel coordinates and computes vehicle speed from the geometric cross-ratio—a perspective-invariant distance ratio—combined with known wheelbase specifications. Validation across three settings demonstrated high accuracy: simulations on straight and curved roads yielded maximum errors of 0.30 km/h and 1.70 km/h, respectively; real vehicle tests achieved errors of 0.41 km/h under constant-speed driving and 0.57 km/h during acceleration; and actual accident case studies produced errors of 0.49–3.05 km/h compared with field measurements. Unlike existing approaches that provide only point or interval estimates, the proposed framework generates continuous speed profiles throughout the accident sequence, enabling dynamic analysis of driver behavior such as acceleration, deceleration, and braking response. When vehicle specifications are available, this approach delivers legally defensible speed estimates from readily available video evidence, significantly improving the precision and practical utility of forensic accident reconstruction.