基于智能视觉的船舶碰撞风险评估模型

A model of ship collision risk assessment based on intelligent vision

构建基于智能视觉的船舶碰撞风险评估模型,以提升船舶航行安全性。利用机器视觉技术采集的船舶航行图像,提取船舶航行图像中的颜色特征与梯度方向特征。利用高效卷积运算符(efficient convolution operators for tracking,ECO)算法学习滤波器模板,获取船舶航行图像特征的响应图值。加权求和处理特征响应图值,实现船舶航行轨迹跟踪。依据船舶航行轨迹跟踪结果的时间与坐标信息,获取船舶航行的速度、方位以及坐标信息。依据船舶航行速度等信息,构建船舶碰撞风险评估模型,输出船舶碰撞风险值。测试结果表明,该模型有效评估船舶航行过程中的对遇碰撞、交叉碰撞以及追越碰撞风险,提升船舶航行安全性。

Build a ship collision risk assessment model based on intelligent vision to improve the safety of ship navigation. Using the ship navigation image collected by machine vision technology, the color features and gradient direction features in the ship navigation image are extracted. The efficient convolution operators for tracking(ECO) algorithm is used to learn the filter template and obtain the response graph value of ship navigation image features. Weighted summation is used to process the value of the characteristic response diagram to realize the tracking of the ship’s navigation trajectory. According to the time and coordinate information of ship trajectory tracking results, the speed, orientation and coordinate information of ship navigation are obtained. According to the information of ship’s sailing speed, the ship collision risk assessment model is constructed, and the ship collision risk value is output. The test results show that the model can effectively evaluate the risks of collision, cross collision and overtaking collision in the process of ship navigation, and improve the safety of ship navigation.