Dataset Used The Vehicle Color Recognition Dataset contains vehicle images in eight colors, which are black, blue, cyan, gray, green, red, white and yellow. The images are taken in the frontal view captured by a high-definition camera with the resolution of × on the urban road. The collected data set is very challenging due to the noise caused by illumination variation, haze, and over exposure. The dataset is available at -

Classifiers The haze free images are fed to a convolutional network for feature engineering and the output is fed to a fully connected network for classification. Multiple color spaces (RGB, HSV and CIE LAB) are tested to determine the color space which is easiest to interpret and leads to higher performance. The final model, when given an image of a vehicle, will be able to predict the color of the vehicle with high accuracy.

Problem Statement In recent years the number of cars on roads has increased exponentially and, identifying them has become a very big task. Vehicle color information is one of the important elements in ITS (Intelligent Traffic System), the other two being– Make and model of the car and license plate recognition. Vehicle color is an important property for vehicle identification and provides visual cues for fast action law enforcement. Recognizing the color of a moving or even a still vehicle can be a very challenging task because of several factors including weather conditions, quality of video/image acquisition, and strip combination of the vehicle. Different ensemble algorithms can be used to give us color recognition.

The color condition of the car is one of the main parameters in buying and selling used cars; This can increase or decrease its price by millions of tomans. Perhaps the best way is to go to a car paint expert who can tell you easily and in full detail with the help of special devices whether the used car in question has been painted or not. In fact, the diagnosis of these people is completely justified and makes the transaction easy for both sides. "Thickness" and "density" are two important characteristics of car paint, knowing them we can easily check the condition of a car&#;s body painting. Pourmohammadi color detection in Najafabad is ready to provide services to you dear ones. Services provided in this center: . Car painting visit. . Loss of chassis. . Inspection of replacement parts. . Presentation of the visit guarantee sheet along with the color chart of the car in a computerized and printed form. Address: Najafabad, East Shariati Street, in front of the water department on Resalat Street

Dehazing of images A major part of this project will deal with the removal of haze from images. This algorithm can be applied to many other problems too. This algorithm uses atmospheric light to give us clearer images. As cities grow bigger, we are encountering more air and light pollution. Both contribute to hazy images that are difficult to process in real time. We will be using transmission maps and atmospheric light to generate a clear image.