一种智轨列车的全景环视图像生成方法

Panoramic surround-view image generation method for ART

针对新型交通系统多编组智轨列车(Autonomous rail rapid transit,ART)由于车体过长而存在的安全隐患,基于8个单目相机传感器和2个铰接角度传感器,提出改进的图像配准方法以生成单车厢环视图像。结合图像映射原理与几何对齐技术融合铰接传感器角度数据实现整车全景环视图像的生成。研究结果表明:在单车厢环视图像生成过程中,改进的基于固定棋盘格内角点的图像配准方法配准精度与效率更高,且实际操作更简单;在多车厢环视图像生成过程中,本文所提出的方法被证明了在不同的铰接角时都能够有效且清晰地生成整车的全景环视图像,这对可视化大型多编组车辆视野盲区以及主动为驾驶员提供行车环境信息从而保证车辆安全驾驶具有重要意义。

The new transportation system autonomous rail rapid transit(ART) has potential safety hazards due to the long body. Panoramic surround-view image generation method for ART was proposed based on eight monocular camera sensors and two hinged angle sensors. An improved image registration method is proposed to generate a single-carriage surround-view image. Then the articulated angle sensor data was fused to realize the panoramic surround-view image of ART according to the principle of image mapping and geometric alignment technology. The experimental results show that the improved image registration method based on the fixed chessboard inner corner points was more accurate and efficient, and the operation is simpler. The proposed method has been proved to be able to effectively and clearly generate panoramic surround-view images of ART at different hinge angles. This is significant for visualizing blind spots for large multi-marshalling vehicles and actively providing driving environment information to the drivers to ensure safety of the vehicle.