A visual distance detection method based on image processing

Abstract

The invention relates to a visualized distance detection method based on image processing. The method includes the steps of acquiring a common image under a current scene and obtaining the total number of edge pixel points, obtaining the total number of infrared thermal image edge pixel points under corresponding scene and corresponding infrared thermal image correction coefficient, establishing a visibility model, and calculating the visibility under current scene according to the visibility model, obtaining corresponding visualized distance by comparing with a stored visibility lookup table, wherein the visibility lookup table contains mapping relations between standard visibility value and visualized distance under different scenes. Compared with the prior art, the method is simple, has lower requirements on image quality and ensures higher detection precision.

Description

technical field [0001] The invention relates to a visual distance detection method, in particular to a visual distance detection method based on image processing. Background technique [0002] Driving visual distance on expressway is an important factor affecting traffic safety. Traditional visibility detection indicators are easily affected by lighting conditions, ambient temperature, rain and snow, etc., and it is difficult to accurately reflect the actual visual distance of the driver. Moreover, the visibility meter is expensive and expensive. The erection method is complicated, and it is difficult to deploy and use on a large scale. [0003] The existing low-visibility video detection algorithm research is mainly the pixel contrast algorithm, which adopts the 0.05 human eye visible contrast threshold stipulated by CIE, including: the calculation method based on data correction, the calculation method of visibility meter correction, and the calculation method of camera se...

AI Technical Summary

Problems solved by technology

[0002] The driving visual distance of expressway is an important factor affecting traffic safety. Traditional visibility detection indicators are easily affected by lighting conditions, ambient temperature, rain and snow, etc., and it is difficult to accurately reflect the actual visual distance of the driver. Moreover, the visibility meter is expensive and expensive. The erection method is complicated, and it is difficult to deploy and use on a large scale

[0003] The existing low-visibility video detection algorithm research is mainly pixel contrast algorithm, which adopts the 0.05 human eye visible contrast threshold specified by CIE, including: calculation method based on data correction, calculation method of visibility meter correction, and calculation method of camera self-calibration; In addition, there are methods of visibility detection based on the brightness of the road surface. These methods are based on the analysis of the ordinary image itself, so the image quality is relatively high. When the quality of the acquired image is low, there is no research significance. At the same time, the proper processing of images and certain deficiencies of related hardware circuits (such as hardware delay, etc.) are also bottlenecks of this type of method.

Images