中文摘要：

野外机器人视觉导航是复杂场景理解一个重要的研究领域，在军事和智能交通等都有广泛的应用。本文提出一种利用光的偏振特性的环境积水区域目标检测方法，利用水体的高偏振度和水体区域具有相似偏振相位的特点，对图像积水区域进行分割。实验结果表明，该方法在复杂自然环境下对积水区域和泥泞区域检测具有良好的效果。与传统机器视觉方法相比较，偏振视觉手段计算简单、准确，与现有图像处理技术结合使用，更具普适性。

英文摘要：

Field robot vision navigation is an important study field in complex scene, and widely used in military and intelligent traffic, etc. A new method of water hazards detection in the field environment is put forward based on polarization characteristic of light. It takes advantage of the physical principle that the light reflected from water surface is partially linearly polarized and the polarization phases of them are more similar than those reflected from the scenes around, and water hazard region of image is divided. The experimental results show that this method is suitable for water hazards scope detection and muddy region detection in complex natural environment. Compared with traditional robot vision method, polarization vision method has advantages of simple calculation, high accuracy, and it is widely used combined with existing image technology.