CNIPA.AI
返回搜索
档案

SYSTEM AND METHOD FOR EXTRINSIC PARAMETER CALIBRATION AND SELF-INSPECTION OF ON-VEHICLE CAMERAS

发明专利审中
2浏览
20权利要求 · 2 独立
§ Ⅰ

卷宗概要

发明人

Chao Wang; Zhaowei Xin; Donghui Wu

IPC 分类

G6T 7/80G6T 5/40G6T 7/13G6V 10/44G6V 20/56

CPC 分类

G6T7/80G6T5/40G6T7/13G6V10/44G6V20/588

The disclosure provides a system and method for calibrating extrinsic parameters of a camera mounted on a vehicle. While the vehicle is parked, the camera captures an image of a plurality of feature points with known world coordinates on a ground plane. The system determines pixel coordinates of the plurality of feature points in an image plane and estimates a homography matrix between the ground plane and the image plane based on a subset of the plurality of feature points. The system generates an initial estimate of the extrinsic parameters based on the estimated homography matrix and applies an optimization technique to obtain calibrated extrinsic parameters based on the initial estimate, the world coordinates, and the pixel coordinates of the plurality of feature points. Applying the optimization technique can include minimizing inverse-projection errors of the plurality of feature points from the image plane to the ground plane.

原文(中文)

The disclosure provides a system and method for calibrating extrinsic parameters of a camera mounted on a vehicle. While the vehicle is parked, the camera captures an image of a plurality of feature points with known world coordinates on a ground plane. The system determines pixel coordinates of the plurality of feature points in an image plane and estimates a homography matrix between the ground plane and the image plane based on a subset of the plurality of feature points. The system generates an initial estimate of the extrinsic parameters based on the estimated homography matrix and applies an optimization technique to obtain calibrated extrinsic parameters based on the initial estimate, the world coordinates, and the pixel coordinates of the plurality of feature points. Applying the optimization technique can include minimizing inverse-projection errors of the plurality of feature points from the image plane to the ground plane.