Fast surveillance video haze removal method

Abstract

The invention discloses a fast surveillance video haze removal method, aiming at addressing high computing complexity and fuzzy details of haze removed images in the prior art. The technical solution includes the following steps: upon computing an atmospheric light, halving a haze image into an upper half and a lower half, obtaining the atmospheric light from the upper half of the haze image by using dark channel prior; in the process of computing the transmittance of a single frame haze image, estimating the transmittance by using the improved fast transmittance estimation method which introduces adaptive parameters; in the process of processing a video, dividing the video into a background and foreground varying part by using the adjacent frame difference method, only computing the transmittance of the foreground varying part, and making the transmittance computing result at the same position of a previous frame as the transmittance of the background part of the current frame. According to the invention, the method herein not only reduces the complexity in computing transmittance, and increases the clarity of details of the haze removal image. The method can be applied to surveillance situations which have complex moving targets and large moving displacement.

Description

technical field [0001] The invention belongs to the field of video and image processing, and relates to a method for quickly defogging monitoring video, which can be used for monitoring occasions that contain complex moving objects and have relatively large motion displacements. Background technique [0002] In foggy weather, the image captured by the imaging device often has image degradation problems such as color shift, contrast drop, and overall whitening, which interferes with the normal work of the outdoor video surveillance system and reduces the system's performance in target detection, The effectiveness that can be exerted in computer vision applications such as tracking and recognition. Therefore, the research on surveillance video defogging has strong practical significance. [0003] At present, the image and video dehazing algorithm based on the atmospheric scattering model is a research hotspot, and the key problem is to solve the atmospheric light and transmit...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to address the shortcomings of the above-mentioned prior art, and propose a method for fast defogging of surveillance video, to reduce the complexity of the transmittance estimation process, improve the video processing speed and the clarity of the image after defogging, and meet the requirements of fast defogging of surveillance video. Fog needs

Images