Vehicle color identification method based on edge detection and convolutional neural network

The invention discloses a vehicle color identification method based on edge detection and a convolutional neural network. The method comprises the following steps: carrying out defogging preprocessing on a vehicle image; performing edge detection and morphological operation on the defogged vehicle image to obtain a binary mask image of an effective color area of the vehicle; constructing and training a color recognition convolutional neural network; and inputting the defogged vehicle image and the corresponding binary mask image, and outputting a color category corresponding to the image. According to the method, the effective color area in the vehicle image is extracted through an edge detection method, the global feature of the whole image and the local feature of the effective color area are integrated, the vehicle color recognition effect is improved, and the effectiveness of the method is verified in an actual traffic checkpoint scene. 本发明公开了一种基于边缘检测和卷积神经网络的车辆颜色识别方法，该方法步骤如下：对车辆图像进行去雾预处理；对去雾后的车辆图像进行边缘检测和形