Face image super-resolution reconstruction method based on two-dimensional multi-set partial least squares

Abstract

The invention discloses a face image super-resolution reconstruction method based on two-dimensional multi-set partial least squares, and the method comprises the following steps: 1, dividing a face image into a plurality of overlapped image blocks, and extracting the potential features of the two-dimensional face image blocks through a two-dimensional multi-set partial least squares method; 2, performing high-frequency face image reconstruction on an input low-resolution face image by using a neighborhood reconstruction strategy and image block combination; and 3, obtaining the finally outputsuper-resolution reconstructed image by adding an input low-resolution face image to the reconstructed high-frequency image. The method has a certain theoretical basis; the two-dimensional multi-setpartial least squares method is innovatively provided, test results under different databases show that the method has high robustness and certain market implementation feasibility, the multi-view super-resolution problem which cannot be processed by most of existing algorithms at present is solved, and the method has high innovativeness and practicability.

Description

technical field [0001] The invention relates to the field of super-resolution reconstruction, in particular to a method for super-resolution reconstruction of face images based on two-dimensional multi-set partial least squares. Background technique [0002] Traditional face recognition methods work under ideal posture and lighting conditions. However, most of the face images captured in real life are low-resolution. In order to solve this problem, many effective face super-resolution reconstruction methods have been proposed in recent years, the goal of which is to reconstruct a high-resolution face image from an input low-resolution image. However, the traditional face super-resolution algorithm cannot maintain the two-dimensional structure of the image, and cannot handle the input of multiple resolution face images at the same time. [0003] In addition to the low-resolution problem, in real life, people usually need to face the situation that the same face has multiple ...

AI Technical Summary

Problems solved by technology

For a variety of different low-resolution views, most current methods can only train a pair of high- and low-resolution views at a time, which is time-consuming and inefficient, and cannot effectively utilize the correlation between different resolution views. Co-learning the relationship between multiple views has not received much attention so far.

Images