Hazy Video denoising using an enHanced Visibility RestoRation tecHnique Anju J Prakash * and A Ferdinand Christopher ** Abstract: The visibility of outdoor videos captured in some weather conditions is often degraded due to the presence of haze, fog, sandstorms. Poor visibility causes failure in computer vision applications, object recognition systems, and intelligent transportation systems as well as in traffic analysis. In order to solve this problem different dehazing techniques are employed. This paper proposes a simple video dehazing method that uses a combination of four major techniques: A dark channel prior technique, A refined transmission technique using guided interpolated filter, and an enhanced transmission technique. The proposed refined transmission takes advantage of the guided interpolation filter technique as it is good working with video and adopts adaptive gamma correction technique for enhanced transmission. By doing so, halo effects can be avoided from videos and effective transmission map estimation can be obtained. Keywords: Hazy image, Bad weather, Depth estimation, Visibility restoration. intRoduction Their arises many difficulties while processing outdoor videos in the presence of haze, fog or smoke which diminishes the colors and reduces the contrast. Many noise removal approaches have been proposed to restore the visibility of faded videos in order to improve system performance in all weather conditions. These approaches can be further divided into 3 categories. They are ∑ Additional information approaches ∑ Multiple image approaches ∑ Single-image approaches. Additional information approaches refine hazy images by using scene depth information obtained from additional operations or interactions, such as through user operation to control position of the camera via a given approximation 3-D geometrical model. However, these approaches J. Kopf et. al., [10] are not well suited for real world assumption due to limitations placed on the acquisition of scene depth information by unknown geography information and additional user operation. Multiple image approaches Schechner et. al., [8] adopt two or more images of the same scene, which are captured by using specific hardware, example; a rotating polarizing filter, to effectively construct the scene depth information and further achieve visibility restoration of incoming hazy images. Unfortunately, the * Assistant Professor, Dept. of Computer Science and Engineering, Sree Buddha College of Engineering, Ayathil, Elavumthitta, Kerala. Email: jpanju@gmail.com ** Assistant Professor, Dept. of Computer Science and Engineering, Noorul Islam University, Kumaracoil, Thuckalay. Email: afchristopher@gmail. com © Serials Publications Man In India, 97 (2) : 29-36