Vehicle Distance Estimation of Monocular Camera based on 3D Bounding Box Detection
- 提供方法
- 本サイト上にてダウンロード・閲覧可
- 形態
- 価格
- 一般価格(税込):¥1,100 会員価格(税込):¥880
- 文献番号
- 20219004
- 文献・情報種別
- その他の国際会議
- 掲載ページ
- 1-3(Total 3 p)
- 発行年月
- 2021年 9月
- 出版社
- (公社)自動車技術会
- 言語
- 英語
書誌事項
著者(英) | 1) Qi Xu, 2) Xingguo Zhang, 3) Pongsathorn Raksincharoensak, 4) Hiroshi Mouri |
---|---|
勤務先(英) | 1) Tokyo University of Agriculture and Techology, 2) Tokyo University of Agriculture and Techology, 3) Tokyo University of Agriculture and Techology, 4) Tokyo University of Agriculture and Techology |
抄録(英) | Recently, the monocular camera’s distance calculation is the key to analyze the near-miss driving recorder data which can be used in automatic driving in the future. This paper presents an accurate method of distance calculation aiming at the image’s feature point found. We use a 3D object tracking method by the results from the 3D bounding box detection method which uses the 2D bounding box to estimate the 3D bounding box. We adopt pitch angle estimation which also is used in our previous research to calculate camera pitch angle and distance. The driving recorder data which is near-miss video is able to detect the correct angle of the 3D bounding box by calculating the pitch angle of each frame in advance for distance calculation. 翻訳 |