Color image acquisition method

Abstract

The invention discloses a color image acquisition method. The method comprises the following steps: firstly, reconstructing a brightness image Y which has a relatively high resolution and can reflect more details under white light; reconstructing a red monochrome image R and a blue monochrome image B which have relatively low resolutions under red light and blue light respectively; calculating a green monochrome image G by using a linear relation according to the acquired brightness image Y, the red monochrome image R and the blue monochrome image B; performing linear conversion according to the obtained red monochrome image, green monochrome image and blue monochrome image to obtain a monochrome image U and a monochrome image V presenting chrominance information; extending the resolutions of the monochrome image U and the monochrome image V to a solution being the same as that of the brightness image Y through up-sampling; and lastly, fusing a component Y, a component U and a component V to reconstruct a color image. Through adoption of the color image acquisition method, the sampling times are reduced effectively; important information is not lost; calculation overhead needed by a CS (Chirp Scaling) algorithm is avoided; and time needed by reconstruction is shortened.

Description

technical field [0001] The invention belongs to color self-adaptive compression computing ghost imaging technology, in particular to a novel color image acquisition method. Background technique [0002] In the past research, computational ghost imaging was mainly used to reconstruct grayscale images, and there were few studies on the acquisition methods of color images. Considering color imaging, a direct and obvious method is to measure and reconstruct the three components of red (R), green (G), and blue (B) separately, and then fuse the three components. Welsh et al. ([1] S S. Welsh, MP. Edgar, R. Bowman, P. Jonathan, B. Sun, and M J. Padgett, "Fast full-colorcomputational imaging with single-pixel detectors," Opt.Express 21(20), 23068-23074(2013).) proposed a color image reconstruction method and system, which uses a digital light projector (DLP) to project speckle to the object under white light, and the reflected light intensity It is divided into three outputs of red...

AI Technical Summary

Problems solved by technology

[0003] As shown by the above two algorithms, the existing algorithms use compressed sensing to reconstruct color images, which require huge computational overhead, the reconstruction time increases exponentially with the imaging resolution, and the imaging speed is slow

In addition, the existing methods are to reconstruct the three components of red, green and blue respectively, which requires more sampling times and poor image quality

Images