An Improved Method for Constructing Color Background of Grayscale Image

Abstract

The invention discloses an improved method for constructing a color background of a gray-scale image, and obtains a panoramic image and a gray-scale scene image through an image splicing technique. Then zoom the grayscale scene graph to the same size as the same position of the panorama image through pixel-level scaling, and then complete the template matching; then perform image inpainting on the movable objects in the grayscale scene graph to obtain the movable objects in the grayscale scene graph. A mask for moving objects. Finally, the mask of the movable object in the grayscale image is combined with the color background image through the image fusion technology based on wavelet transform, and finally the color background image containing the movable object in the grayscale scene image is obtained. The present invention uniquely considers the movable object information contained in the grayscale scene graph itself, and extracts the movable objects that can be used as the background and puts them into the color background graph obtained by template matching. The invention adopts the image fusion technology based on wavelet transform, so that the effect of fusion can be used in work.

Description

technical field [0001] The invention relates to image processing technology, in particular to an improved method for constructing a color background of a grayscale image. Background technique [0002] With the ever-changing computer technology and the urgent need of the Huludao project, it is necessary to be able to use the infrared LED camera to shoot the target scene in a series of bad weather such as night fog, but also want to obtain the color background information of the scene in the daytime, In this way, more information than infrared video can be obtained in fields such as video surveillance. [0003] The applicant previously proposed a method for constructing a color background of a grayscale target image (application number CN201810661892.8), which uses an infrared camera to capture multiple color scene images; obtains a panorama of a fixed area through image stitching technology; and then Use an infrared camera to obtain a grayscale scene map of a certain part of...

AI Technical Summary

Problems solved by technology

[0004] However, the biggest limitation of this method of constructing a colored background for a grayscale image is that it can only provide background information based on the previously obtained panorama, and cannot take into account the changes brought about by the grayscale image itself, such as when the night is foggy. The obtained grayscale scene image contains movable objects, such as cars, and this object exists as a background object for a period of time at that time, and people cannot integrate this background object into the color background image as a background target only through template matching Go, and people don't want to just paste the content of the rectangular frame obtained by object recognition into the color background image as the background, but hope to completely subtract the movable object along the outline of the movable object and then blend it into the background in the picture

Images