An Improved Dark Channel Prior for Fast Dehazing of Outdoor Images

Object detection has a vital role in many of the real time out door computer vision-based applications. For these applications haze removal become a pre-processing technique, that can recover clear images from foggy ones, which is necessary for object detection. An efficient fast haze removal method based on the atmosphere scattering model and the dark channel prior method is proposed. Instead of using single global atmospheric light to restore foggy image, a local atmospheric light estimation method is applied in the proposed design to achieve optimal results. The existing DCP based methods are time consuming as the transmission map estimation can be executed only after the atmospheric light estimation. The proposed methods exploit the depth map of a image for the transmission map estimation, which results in much more faster dehazing algorithm. The validity of the proposed algorithm is verified using the databases with (O-HAZY) and without ground truth information. The performance of the algorithm is analyzed on a pair of ground truth and hazy images whose PSNR, RMSE have been evaluated and analogized with the existing algorithms. The performance metrics of the proposed algorithm are found to be improved than the other exsisting algorithms.