Using Raindrops Removal Algorithm to Improve Vehicle Recognition via AttentiveGAN

When the vehicle image with raindrops is converted into a clear image, the biggest difficulty is that the area blocked by the raindrops is random, and the entire image loses many feature points. Therefore, the method of removing raindrops from a single image via Deep-Learning can effectively extract and mine the depth features in the image. We propose to use AttentiveGAN to remove raindrops from images, which is based on Generative Adversarial Networks (GAN). Using Generator and Discriminator in the GAN, raindrops in the vehicle images are better removed, thereby effectively improving the recognition rate of vehicles. From the final experimental evaluation, the images generated via this algorithm have greatly improved the recognition appearance, PSNR and SSIM, and the recognition rate of the vehicles can be increased by up to 30%.