Video slight change amplification method based on PCA

Abstract

The invention discloses a method for amplifying small change of video based on PCA, The implementation process is as follows: first converting the input video image from the RGB color space to the YIQcolor space, each frame of the video is then decomposed into spatial Gaussian/Laplacian pyramids in a frame cycle manner, and then PCA processing is carried out on the data decomposed by spatial scale, to get signals of small changes of interest, then by directly multiplying the magnification factor set by one person, Realize the enhancement of the small signal amplitude, then reconstruct the Gaussian/Laplace pyramid, and finally add it with the original video source, and transform the processed video image from YIQ color space to RGB color space, and obtain the enlarged result of the small change of the video. As the PCA processing itself has the function of suppressing noise, the method of the invention can effectively suppress noise while amplifying a slight change, and the practical application of the video amplification method is expanded.

Description

technical field [0001] The invention belongs to the field of video image processing, and in particular relates to a PCA-based method for amplifying small video changes. Background technique [0002] The human visual system has its sensitivity limitation in the spatial domain, and the naked eye cannot perceive changes beyond the perception domain. However, many signals beyond the human visual ability are highly informative. For example, blood circulation causes subtle periodic changes in the skin of the human body. This change, which cannot be perceived by the naked eye, is very consistent with the human heart rate. The small-amplitude vibrations in the spatial domain of the wrist are difficult to see with the naked eye, but they can be clearly perceived after magnification, thus revealing interesting physiological characteristics of the human body. The tiny vibrations of mechanical grinding tools, the slight shaking of bridges and buildings, etc., cannot be directly percei...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem that the linear Euler video amplification technology is sensitive to noise, and proposes a PCA-based method for amplifying small changes in video, and uses PCA processing in combination with Euler perspective to amplify the small changes in the video under noise conditions. It can be used in health and medical fields such as heart rate detection, pulse wave extraction and detection, and can also be used in engineering applications such as monitoring bridge vibration, mechanical vibration, and abrasive tool quality inspection; solves the problem of traditional linear EVM amplifying color signals or motion signals while amplifying noise problem, improve the anti-noise performance of video small change amplification method, and further expand the application of video amplification technology in actual scenes

Images