该课题为基于MATLAB bp神经网络的雾霾天气下交通标志的识别系统。主要分两步骤，一是进行图像去雾，采用暗通道的方法获取光透射率，从而去除雾霾。得到清晰的图片后，利用颜色的方法进行交通标志的定位，众所周知，交通标志基本是红，蓝，黄三色组成，根据RGB不同组合可以定位到不同颜色，因为存在误差，所以需要借助形态学相关知识，将得到的误干扰面积去除，从而实现精准定位。定位后，在原图基础上进行分割出彩色图标，利用bp神经网络方法，进行训练，识别，从而得出结果。本设计配有一个GUI可视化界面，操作简单容易上手。是个不错的选题。(This project is a traffic sign recognition system based on Matlab bp neural network in haze weather. There are two steps. One is image defogging, and the dark channel method is used to obtain light transmittance to remove haze. After getting clear pictures, use color method to locate traffic signs. As we all know, traffic signs are basically composed of red, blue and yellow. According to different combinations of RGB, different colors can be located. Because there are errors, we need to use morphological knowledge to remove the error interference area, so as to achieve accurate positioning. After positioning, the color icon is segmented on the basis of the original image. The BP neural network method is used to train and identify the color icon, and the result is obtained. This design is equipped with a GUI visual interface, which is easy to operate. It's a good topic.)