Image separation method and device, computer device and storage medium

Abstract

The embodiment of the invention discloses an image separation method and device, a computer device and a storage medium. The method comprises the steps of: acquiring a foreground pixel value and a background pixel value for separating a target selection area from an original image; calculating foreground geodesic distances from respective pixels in the original image to the foreground pixel value,and background geodetic distances from respective pixels in the original image to the background pixel value; calculating the transparency of the target selection area according to the foreground geodesic distances and the background geodesic distances to separate the target selection area from the original image based on the transparency. During the image separation, by giving the foreground pixel value and the background pixel value for extracting the target selection area and then calculating the transparency of each pixel by the shortest spatial distance between the pixels and the foreground pixel value and the background pixel value, the contour extraction of the selection area is made accurate.

Description

technical field [0001] Embodiments of the present invention relate to the field of model algorithms, in particular, an image separation method, device, computer equipment and storage medium. Background technique [0002] "Cutting out" is one of the most commonly done operations in image processing, which refers to extracting the required part of the image from the screen precisely. Image matting is an important basis for subsequent image processing. A large number of matting techniques are used in movies and film and television dramas to synthesize fantasy images. Channel matting method is a common and most widely used matting method among many matting methods. [0003] In the prior art, the principle of channel deduction is as follows: channels include: color channel, ALPHA channel, and spot color channel. Among them, the ALPHA channel is used to store the selected area, and the ALPHA channel uses the 8-bit grayscale between black and white to save the selection. Instead...

AI Technical Summary

Problems solved by technology

[0004] The inventors of the present invention found in the research that, in the prior art, when the channel is cut out, it is necessary to manually repair or delete the edge of the selected range when selecting the range, otherwise, the edge of the determined selected range is blunt, and the cutout effect is not natural enough. Unable to naturally integrate the selected image into the new image environment

Images