Outdoor image and video corrupt easily owing to fog, mist etc. Several techniques have been proposed to overcome these problems. But there is no optimum method as these scenarios are unpredictable and all situations are unique in their own way. Dark channel prior — a new prior-gives a significant prediction about the quantity of airlight that causes haze. Dark channel with contrast controller and sky masking improves the visibility with significantly reduced time. In this paper a innovative technique has been studied and it's qualitative and quantitative analysis carried out.

Dehazing technique based on dark channel prior model with sky masking and its quantitative analysis

Visibility Improvement is a great challenge in early vision. Numerous methods have been experimented. As the subject is random and different significant parameters are involved to improve the vision, it becomes difficult, sometimes unsolvable. In the process original image has to be retrieved back from a degraded version of the image which is often difficult to perceive. Thus the problem becomes ill-posed Inverse Problem. This has been observed that VI (Visibility Improvement) is associated with haze and blur. This complex nature requires probability distribution, estimation, airlight calculation etc. In this paper a combination of haze and blur model has been proposed with detail discussions.

http://www.mecs-press.org/ijmecs/ijmecs-v8-n12/IJMECS-V8-N12-7.pdf

Modeling of Haze Image as Ill-Posed Inverse Problem & its Solution

At present the classical problem of visibility improvement is hot topic of research. An image formation optical model is presented where a clear day image has high contrast with respect to an image plagued with bad weather. A degraded daytime image has high intensity with minimum deviation among pixels in every channel. No reference digital image haze removal is a problem. The static haziness factor for all types of images cannot be applicable for effective haze removal. A minimum intensity channel of the three RGB channels is estimated as transmission of an image with a dynamic haziness factor to be a ratio of minimum to maximum pixel intensity of the hazy image. Adaptive contrast, extinction coefficient, the maximum visible distance of hazy images as well as dehazed images from each image are evaluated uniquely. The resulting high-quality haze free image with linear computational complexity O(n) is appropriate for real time applications. The effectiveness of the technique is validated by quantitative, and qualitative evaluations.

Fast Single Image Haze Removal Scheme Using Self-Adjusting: Haziness Factor Evaluation

The invention discloses an image defogging method, a storage medium and a terminal device, which are used to solve the problem on how to restore an image containing a bright sky area into a fog-free image. The method provided by the invention includes the following steps: acquiring a target image to be defogged and a gray image thereof; calculating a gradient map, and calculating a sky mask map according to the gradient map and a preset gradient threshold; calculating the mean neighborhood brightness of the target pixels in the gray image; calculating a sky weight map according to the mean neighborhood brightness of the target pixels and a brightness threshold; calculating a dark channel map of the target image; correcting the dark channel map and filtering the dark channel map in a guided manner according to the sky weight map; determining the atmospheric light value of the target image according to the distribution of the mask points in the sky mask map; calculating a transmission map according to the atmospheric light value, a preset defogging intensity factor and the filtered dark channel map; and getting a defogged image through calculation according to the transmission map and an atmospheric scattering model.

Image defogging method, storage medium and terminal device

In this paper, the image haze-removal technique proposed by us is free from artifacts as well as it's execution time is also very less. The method proposed here can overcome the halo-artifacts occurring due to non-uniform transmission within patches by patch size regulation and is also capable of removing the blocking artifacts (especially in the sky region) and can restore the contrast of the image by using sigmoid function and fuzzy logic. Moreover, since we have taken a large number of hazy images in our database and have obtained satisfactory results by applying the algorithm on all the hazy images, hence we can state that our algorithm is not dependent on the features of the hazy images and can remove haze from all kinds of hazy images irrespective of its characteristics.

Fuzzy logic based vision enhancement using sigmoid function

Images and videos of outdoor scenes suffer from reduced clarity due to presence of fog/haze/mist, and thus it becomes difficult to drive in bad weather conditions. Several methods have already been proposed to improve the images acquired in foggy weather conditions. In this paper a novel method of dehazing using dark channel prior along with masking the sky regions has been proposed, the output has improved considerably due to clear visibility of separation of surrounding edges from the sky as well as reduced artifacts. Focus on road edge detection has also been emphasized on, in this work along with dehazing leading to prominent visibility in foggy conditions.

Dibyasree Das

Papers

Dehazing technique based on dark channel prior model with sky masking and its quantitative analysis

Visibility Enhancement in a Foggy Road Along with Road Boundary Detection