Method for estimating visibility of single fog-degraded image

Abstract

The invention provides a method for estimating visibility of a single fog-degraded image on the basis of scene depth. The visibility serves as a basic element of meteorological observation and is an important evaluation index in the field of fog-degraded image processing. The traditional visibility testing method depends on hardware facilities such as atmospheric transmission instruments, and multi-point manual observation is required. The method is used for estimating transmissivity on the basis of an atmospheric degradation physical model, the scene depth is estimated by a scenery depth estimation method, and an atmospheric extinction coefficient is solved according to a principle of a visibility measuring instrument. Probability statistic thinking is adopted, the visibility on a single pixel point is estimated, the probability of the visibility of pixels of different grades is counted according to the grade of the visibility in foggy days, and the visibility of the single fog-degraded image is estimated. By the method, only the single fog-degraded image is estimated, calculation is simple, the estimation accuracy is high, hardware facilities are not required, and manual operations are reduced. The method can be used for effectively evaluating a defogging effect in a foggy day image sharpening system.

Description

Technical field: [0001] The present invention relates to an objective evaluation standard for picture quality in image / video processing technology, and specifically relates to a method for estimating the visibility of a single foggy image based on scene depth. technical background: [0002] The clear processing of foggy degraded images is one of the research hotspots in the field of computational vision and imagery, and has broad application prospects. The relevant research results can be widely used in aerial photography, video surveillance, photo processing, and safety assistance for various vehicles. Driving systems and many other fields. However, at present, there is no unified comprehensive evaluation system for the performance evaluation of the dehazing algorithm, and most researchers use their own evaluation indicators to measure the performance of the proposed algorithm, resulting in a lack of good comparability between algorithms. There are three main reasons why t...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the limitations of existing visibility testing methods that rely on hardware facilities such as atmospheric transmissometers and require manual multi-point observation, and provide a simple and accurate visibility estimation based on a single image based on transmittance estimation and scene depth method, by which the visibility of foggy images can be easily calculated, and a standard is provided for evaluating the performance of the dehazing algorithm

Images