Image automatic focusing method and device, apparatus and readable storage medium

Abstract

The embodiment of the invention discloses an image automatic focusing method and device, an apparatus and a computer-readable storage medium. The method comprises the following steps: segmenting the original image into a plurality of sub-regions of different target objects to obtain a region segmentation map; inputting the image into the depth-of-field estimation network to obtain the depth-of-field estimation map; dividing the depth-of-field estimation map into a plurality of corresponding sub-regions by using the region partition map, and obtaining the average depth-of-field values of each sub-region; determining a main body portion and a background portion in the original image according to the acquired main body image block and the region segmentation image; according to the differencebetween the average depth of field of each sub-region of the background part and the average depth of field of the main image block, grammaticalizing the background of the virtual background image tobe grammaticalized in different degrees, wherein the grammaticalization degree increases with the increase of the difference between the average depth of field of the main image block and the grammaticalization degree of the background image. The present application can more accurately and naturally highlight the image body part, so that the picture body and the background edge after the background grammaticalization are reasonable, and the transition is natural.

Description

technical field [0001] Embodiments of the present invention relate to the field of image processing, and in particular, to an image automatic focusing method, device, equipment, and computer-readable storage medium. Background technique [0002] With the rapid development of camera and photography technology, human beings are pursuing more and more beautiful things, and image processing technology has also been developed accordingly. Image blurring technology was mainly used in SLR cameras to take pictures at the beginning. It is realized by adjusting various parameters of the camera, such as aperture, shutter speed, etc. The quality of image blurring effect directly depends on the performance of SLR cameras. In addition, image blurring techniques also rely on image editing software, such as image background blurring through cutouts and blurring. With the development of image processing technology, some more intelligent background blurring methods that do not depend on came...

AI Technical Summary

Problems solved by technology

This is more complex and inconvenient, with a lower degree of intelligence and automation

In addition, this method uses the principle of binocular vision to obtain the depth information of the scene, which relies on the high-cost binocular camera, which is not conducive to the use of general users and has poor universality.

Images