This project focuses on calculating the similarity between the images using genetic algorithm and haar-wavelet. Haar wavelet is being used in order to calculate the images with different features like various expressions or gloomy or other characteri

交通信号灯的检测是复杂场景下交通灯识别的重点。采用了色彩分割与关联滤波方案进行交通灯的检测。首先，建立交通信号灯的高斯模型，利用高斯向量与多色彩空间结合的方法进行图像分割。然后，采用基于区域增长与相似性判定的关联滤波，对色彩分割后的图像进行处理。最后，使用基于canny算子的边缘提取算法获取方向指示灯轮廓特征，并使用基于改进hu不变矩和马氏距离对方向指示信号灯进行分类。-The detection of traffic lights is the focus of traffic lights