A fog-penetrating imaging method based on binocular cameras

Abstract

The invention discloses a fog-penetrating camera method based on a binocular camera and relates to the field of camera monitoring. This method is as follows: ① The two cameras of the binocular camera image the scene separately; ② Find out the same scene points in the two cameras by comparing the contours, and calculate the coordinates of each point; ③ Slice the scene according to the distance, the same z a Cut the scene into one slice; ④ execute different defogging algorithms for slices with different distances or use the same defogging algorithm to execute different defogging parameters; ⑤ output images. According to the distance between objects in the scene and the camera, the present invention adjusts different fog penetration algorithms or adopts the same fog penetration algorithm to implement different fog penetration parameters; close-range scene images are relatively clear, so that the fog penetration effect is small and more original images are retained Image information: The effect of fog penetration in distant scenes is strong, and more image noise is removed to make the image clear; in this way, the entire image is clearer and less distorted in the final imaging.

Description

technical field [0001] The invention relates to the field of camera monitoring, in particular to a fog-penetrating camera method based on a binocular camera. Background technique [0002] With more and more hazy weather, the cameras used in actual security projects are also increasingly emphasizing the fog penetration function. At present, there are two main methods of fog penetration for cameras: optical fog penetration and electronic fog penetration. Optical defogging mainly uses infrared rays to better pass through tiny particles to form images to achieve the effect of fog penetration; electronic fog penetration mainly uses algorithms to process images to achieve the effect of fog penetration. [0003] Although the effect of optical fog penetration is relatively good, it needs to use high-cost fog penetration lens, and the volume of fog penetration lens is generally relatively large, so the practical application is limited. [0004] In actual engineering, the electronic...

AI Technical Summary

Problems solved by technology

[0003] Although the effect of optical fog penetration is relatively good, it needs to use high-cost fog penetration lens, and the volume of fog penetration lens is generally relatively large, which limits the practical application

[0004] In actual engineering, the electronic fog penetration method is widely used, but the fog penetration effect is not ideal

Most of the current electronic fog penetration algorithms use the same strength fog penetration algorithm for the entire image to process the fog, so there is a problem: due to the effect of fog, the near-distance scene image is relatively clear, and the far-distance scene image is relatively blurred

If the fog penetration algorithm is designed according to the most blurred or relatively blurred places in the image, the relatively clear close-range scene information will be lost more; if the fog penetration algorithm is designed according to the clearest or relatively clear places in the image, the distance scene The effect of fog penetration is not good

Images