Interactive matting method, medium and computer device

Abstract

The invention discloses an interactive matting method, a medium and a computer device. The foreground sample space and background sample space are obtained by human-computer interaction. According toa marking operation instruction, smearing a hair region of the original image to mark an unknown region; Obtaining pixels of each unknown region, traversing all sample pairs according to each pixel ofthe unknown region to select a sample pair with the smallest total cost function value, and calculating an alpha value corresponding to each pixel of the unknown region according to the selected sample pair; The alpha mask image is obtained according to the alpha value corresponding to each pixel of the unknown region, and the alpha mask image is denoised to obtain the final alpha mask image; andthe alpha mask image is denoised to obtain the final alpha mask image. Thus, sample pairs and unknown regions can be determined through simple interaction, and the alpha value of pixels in the unknown regions can be calculated according to the sample pairs, so that users can complete accurate matting of hair edges without mastering rich PS technology and color channel knowledge.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an interactive map-cutting method, medium and computer equipment. Background technique [0002] Cutout is one of the most common operations in image processing, which refers to the operation process of extracting the required part of the image from the screen. [0003] In the actual process of cutout operation, when encountering the processing of images containing human hair, animal hair, etc., if it is purely manually cut out without tools, each picture will take a lot of time and energy for the operator; therefore , in order to solve the problem of difficult extraction of such images, existing image extraction techniques have proposed sampling methods such as Knockout and RobustMatting to improve the efficiency of operators in extracting target images; however, most of these sampling methods are very complicated and require Operators have a wealth of PS technology and...

AI Technical Summary

Problems solved by technology

[0003] In the actual process of cutout operation, when encountering the processing of images containing human hair, animal hair, etc., if it is purely manually cut out without tools, each picture will take a lot of time and energy for the operator; therefore , in order to solve the problem of difficult extraction of such images, existing image extraction techniques have proposed sampling methods such as Knockout and RobustMatting to improve the efficiency of operators in extracting target images; however, most of these sampling methods are very complicated and require Operators have a wealth of PS technology and color channel knowledge, which is difficult for beginners to operate

Images