Face portrait compositing method based on single photos and portrait pairs

Abstract

The invention discloses a face portrait compositing method based on single photos and portrait pairs. The method comprises randomly selecting a pair of photos from a database sample set as an initial image pair and the others photos as a test photo sample set; dividing image blocks obtained by blocking the initial image to obtain a first image block subset and a second image block subset; performing matrix decomposition on the obtained first image block subset to obtain a projection matrix, and projecting the second image block subset onto the projection matrix to obtain a first projection coefficient set; projecting image blocks obtained by blocking a test photo onto the projection matrix to obtain a second projection coefficient set; according to the similarity scale distance, compositing matched image blocks searched in the second image block subset into an initial composite image; performing iterative composition to obtain a final composite image. The face portrait compositing method based on the single photos and the portrait pairs improves the quality of the initial composite image and saves the process of taking time and labor to collect a large number of photos and portrait pairs.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a face portrait synthesis method based on a single photo and a portrait pair. Background technique [0002] Face portrait synthesis has many important practical applications, such as assisting the police in arresting criminals and digital entertainment. When the criminal's photo cannot be obtained due to certain factors, the police need to draw a portrait of the suspect through the description of the witnesses, and then match it in the police's face photo database. Due to the large difference in the modality of photos and portraits, all photos in the police face photo database must be converted into portraits before matching, and then the suspect's portraits will be matched in the synthetic portrait database. At present, social networks such as Twitter and Weibo are becoming more and more popular, and many users upload their own portraits to the Internet as ...

AI Technical Summary

Problems solved by technology

However, this method has the following disadvantages: each synthetic image block is synthesized separately, and the neighborhood relationship between the image blocks is not considered, resulting in an unnatural image at the fusion point and noise.

This method has the following defects: each test photo block finally selects only a training image block instead of a synthetic image block as the image block to be synthesized for image synthesis, resulting in block effects and missing details in the synthesized image

This method has the following defects: the quality of the synthesized image depends on the initial image synthesis method, and the linear superposition of detail information makes the synthesized image unnatural

[0007] The above three types of image synthesis methods are only applicable to the situation where the background of the test photos and the training set photos are the same. When the test photos and the training set photos have different backgrounds, the background of the test photos synthesized by the above three methods will have noise.

When there are non-face parts such as glasses and hairpins in the test photos, but these non-face parts do not exist in the photos of the training set, the corresponding portraits of the test photos synthesized by the above three methods will lose the non-human parts in the test photos. face parts

In addition, the above three types of portrait synthesis methods all need to use a large number of photo-portrait pairs as the training set. When there is only a single photo-portrait pair in the training set, the synthesis effect is very poor. However, collecting a large number of photo-portrait pairs is time-consuming and laborious.

Images