Used a pre-trained YoloV3 model for detecting vehicles in any given video.
After detecting the vehicle,trained a classifier along with image processing methods in OpenCV for color recognition.
You can download the weights from https://pjreddie.com/darknet/yolo/