What question did this study set out to answer?

This research aims to develop a method for accurately surveying traffic volume using LiDAR data, particularly in the presence of occlusion.

March 29, 2026Japanese Journal of JSCE0 citations

Development of an Occlusion-Robust Vehicle Classification-Based Traffic Volume Survey Method Using Dynamic Point Cloud Data Measured by Lidar

View Full Paper

YYYuhei YamamotoKansai University MNMasaya NAKAHARA RSRyo SUMIYOSHIHosei University

Key Points

This research aims to develop a method for accurately surveying traffic volume using LiDAR data, particularly in the presence of occlusion.
Utilized dynamic point cloud data from LiDAR for vehicle detection.
Analyzed vehicle direction, speed, and dimensions to improve counting accuracy.
Developed a classification method for distinguishing vehicle types amidst occlusion.
Achieved over 95% accuracy in counting traffic volume.
Attained more than 80% accuracy in vehicle classification.

Abstract

近年，道路交通に関する課題解決に向け，効率的な交通データの計測が求められている．とりわけ自動車交通量調査では，動画像をAIで解析する手法が実用化されつつあるが，照度の影響を受ける課題がある．これに対し，照度の影響を受けにくく，対象物を三次元の点群として計測可能なLiDARが注目されている．既存研究では，LiDARで計測した動的な点群データから，断面交通量を計数する手法が考案されているが，LiDARから遠い車線では，オクルージョンが頻発するため，計数が困難である．そこで本研究では，動的な点群データから検出した車両の進行方向や速度，寸法からオクルージョンに頑健な車種別の交通量調査手法を考案した．実証実験の結果，95%以上の精度で断面交通量を計数でき，車種分類においても80%以上の精度を達成した．

Ask AI

Helpful

Bookmark

View Full Paper

Cite This Study

Yamamoto et al. (2026) studied this question.

synapsesocial.com/papers/69c8c195de0f0f753b39bdd1 https://doi.org/10.2208/jscejj.25-22007

Ask AI

Helpful

Bookmark

View Full Paper

March 29, 2026Japanese Journal of JSCE0 citations

Development of an Occlusion-Robust Vehicle Classification-Based Traffic Volume Survey Method Using Dynamic Point Cloud Data Measured by Lidar

View Full Paper

YYYuhei YamamotoKansai University MNMasaya NAKAHARA RSRyo SUMIYOSHIHosei University

Key Points

This research aims to develop a method for accurately surveying traffic volume using LiDAR data, particularly in the presence of occlusion.
Utilized dynamic point cloud data from LiDAR for vehicle detection.
Analyzed vehicle direction, speed, and dimensions to improve counting accuracy.
Developed a classification method for distinguishing vehicle types amidst occlusion.
Achieved over 95% accuracy in counting traffic volume.
Attained more than 80% accuracy in vehicle classification.

Abstract

Ask AI

Helpful

Bookmark

View Full Paper

Cite This Study

Yamamoto et al. (2026) studied this question.

synapsesocial.com/papers/69c8c195de0f0f753b39bdd1 https://doi.org/10.2208/jscejj.25-22007

Ask AI

Helpful

Bookmark

View Full Paper