Image background automatic replacement method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses an image background automatic replacement method and device, equipment and a medium, and the method comprises the steps: carrying out the semantic segmentation of an image to be subjected to background replacement, carrying out the binary separation of a target object and a background in the image to be subjected to background replacement, and obtaining a binary image; processing an interaction area of the target object and the background in the binary image to enable the interaction area to form an undetermined area different from the target object and the background, and obtaining a ternary image; and performing background replacement on the image to be subjected to background replacement based on the ternary image, the image to be subjected to background replacement and the target background image. By automatically extracting the target object pixels in the intercepted target object region picture, the clear target object image is generated to realize background replacement, so that the generated background replacement picture is more accurate in details, and the precision and efficiency of image background replacement are effectively improved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method, device, electronic device and storage medium for automatic image background replacement. Background technique [0002] Portrait photo background replacement technology refers to the use of color image segmentation technology to process photos in portrait photos, extract the portrait and separate the background, and replace the separated background with other colors or other backgrounds. It is widely used in document entry systems and photo update systems. and image processing systems. [0003] The existing mainstream portrait background replacement mainly adopts the method based on convolutional neural network to complete the replacement. However, this portrait background replacement method is sensitive to the image shooting environment and image quality, and has the defect of low robustness, especially for complex For portrait images captured in the s...

AI Technical Summary

Problems solved by technology

[0003] The existing mainstream portrait background replacement mainly uses the method based on convolutional neural network to complete the replacement. However, this portrait background replacement method is sensitive to the image shooting environment and image quality, and has the defect of low robustness, especially for complex Portrait images taken in the scene, this method generates portrait background replacement results are often unsatisfactory

Images