基于视觉图像的水面机器人波浪检测方法

视觉感知能够为水面机器人的控制策略提供环境依据，而检测水面波浪等级是水面机器人视觉感知的一个重要方面。根据水面波浪环境具有不同的纹理特征，基于波浪图像的傅里叶变换，通过对不同等级、不同光照条件下波浪图像的频域特性分析，提出一种根据不同矩形环之间能量比的关系来判断波浪等级的方法，该方法首先对所采集图像进行直方图均衡化处理，以增强图像的波浪特征，然后进行傅里叶变换，在频域内采用矩形环能量比法提取图像的特征，最后根据第一级和第二级矩形环内能量比的关系来判断波浪等级。实验结果表明，该方法可以有效地检测出水面常见波浪的等级，具有受光照条件影响小的优点。

Method of USV detecting wave grade based on visual image

Visual perception can provide environmental information for the control strategy of the unmanned surface vehicle, and the detection of wave grade is an important aspect of visual perception. Based on the theory of Fourier transform, this paper analyzes frequency domain characteristics of the wave image in different light and sea state. According to the relationship between different rectangle rings energy percentage, a method of judging wave grade is proposed. The characteristic of the wave image is enhanced through histogram equalization; The image is transformed to frequency domain by FT; The image features are extracted by the method of rectangle rings energy percentage; Waves grade are judged by the relationship between the first and the second stage rectangle rings energy percentage. The experimental results show that this method can effectively detect the common wave grade and have the advantage that light conditions have less effect on it.