Real-time video defogging method

Abstract

The invention provides a color keeping based real-time video defogging method in order to effectively recover the contrast and color of an image / video in hazy weather in real time. The method comprises the following steps: firstly, using an optimized orientated filtering method to give a parameter self-adaptive selection method; secondly, providing a simple and effective method to correct the situation that dark channel prior is not suitable for the sky and other large-area white regions; providing a self-adaptive brightness adjustment method capable of keeping the color of an object in order to solve the dark problem of the defogged image, so that the image is realer and more natural; finally, in order to process high-resolution images in real time, adopting a downsampling method for a part with improved transmissivity so as to increase the speed. The time complexity of the algorithm provided by the invention is only a linear function of the number of image pixels, the processing speed is quick, and the consumed time is 80ms for an image with resolution of 600 x 400. The method provided by the invention provides a powerful competitive advantage for enhancing the performances of digital and monitoring products, and provides a core competitive advantage for carrying out the development of video processing chips and intelligent surveillance products and other fields.

Description

Technical field: [0001] The invention relates to image / video processing technology, in particular to a color-preserving real-time video defogging algorithm. technical background: [0002] In recent years, smog has occurred frequently. In foggy weather, the visibility is reduced, which seriously affects people's normal life and the normal work of the video system. Improving the clarity and visibility of hazy weather images / videos in real time has important practical significance. It can be widely used in traffic safety, video surveillance and many other fields. [0003] Foggy image clearing technology has been developed so far, and there are currently two main methods: one is an enhancement method based on image processing; the other is a restoration method based on a physical model. [0004] The purpose of image enhancement is to make the restored image more in line with human visual habits or facilitate the recognition of computer systems. The dehazing method based on i...

AI Technical Summary

Problems solved by technology

[0009] The purpose of the present invention is to solve the problems that the existing defogging algorithm has poor applicability to the environment, poor real-time performance, and is difficult to apply to practice. Based on the atmospheric scattering model, a real-time video defogging algorithm that maintains the color of objects is provided.

Images