一种基于视觉的目标舷角实时测量方法

舰船的舷角是鱼雷攻击时的重要参数,同时也是海航避碰中避碰线路确定的计算依据。在基于视觉的识别中,获取目标舰船的舷角即可确定其相对拍摄视角,从而大大缩小在全方位视野特征库中的搜索范围,提高识别速度。然而现有的舷角测量方法均是采用连续跟踪解算实现的,无法实时测量。为解决实时测量的问题,首先推导了基于视觉的目标舰船舷角测量原理,然后针对摄像机倾斜的情况提出基于水天线校正的改进方法,最后通过实验验证,结果表明方法简单易操作、实时性好。

A Method for Target's Bearing Real-time Measurement Based on Visual Sensing

The target's bearing is an important parameter for torpedo attack,and with which it can compute the route in navigation collision avoidance. In vision recognition,obtaining the target's bearing can determine its relative perspective for me,so it can narrow the search scope in the characteristics library which established by Omni-directional vision of targets,thus increasing the recognition speed. However,the existing method for measuring the target's bearing all based on continuous tracking and computing,cannot measure at real-time. In order to resolve this issue,it first deduces the principle of target's bearing measurement based on vision sensing,then it proposed an improvement method based on water-sky-line rectification for the condition of camera tilt. In the end it verified by experiment, result shows that this method is simple and easy operation,good real-time performance.