Method for compressing color computer hologram through quantum neural network based on gradient optimization

Abstract

The invention provides a method for compressing a color computer hologram through a quantum neural network based on gradient optimization, and belongs to the technical field of compression and transmission of computer holograms. According to the invention, a color computer hologram is manufactured by adopting a bandwidth-limited angular spectrum method for calculation and a double-phase encoding mode. A method for compressing a color computer hologram in an RGB color space through a gradient-optimized quantum neural network is provided, and corresponding network models are trained according to the characteristics of computer holograms with different wavelengths. The computer hologram decompressed by the method is reproduced by adopting a bandwidth-limited angular spectrum method, and is synthesized into a color reproduced image. According to the invention, the compression of the color computer hologram is accelerated, and higher decompressed color computer hologram and color reproduced image quality are obtained.

Description

technical field [0001] The invention provides a method for compressing a color computer hologram by a gradient-optimized quantum neural network, and specifically relates to the technical field of compression and transmission of computer holograms. [0002] technical background [0003] The rapid development of computer technology has provided more solutions for the preparation and calculation of computer holograms. However, based on the characteristics that computer holograms are interference fringes and contain a huge amount of original object information, the method of compression transmission is still worth studying. difficult problem. [0004] In 2010, BP neural network was applied to the compression of grayscale computer hologram [1] , the paper proposes a system for hologram compression and transmission, showing that the method using neural networks is more robust than DCT and DWT compression at low compression ratios, but the compression speed of this method and the q...

AI Technical Summary

Problems solved by technology

[0003] The rapid development of computer technology provides more solutions for the preparation and calculation of computer holograms. However, based on the characteristics of computer holograms, which are interference fringes and contain a huge amount of original object information, the method of compression transmission is still worth studying. Difficult question

[0004] In 2010, BP neural network was applied to the compression of grayscale computer holograms [1] , the article proposes a system for hologram compression and transmission, and shows that the method using neural networks is more robust than DCT and DWT compression at low compression ratios, but the compression speed of this method and the quality of reconstructed and reproduced images are still limited

There are still problems in this field: no method for compressing color computer holograms has been proposed, and the convergence speed and convergence effect in the backpropagation process of quantum neural networks need to be further optimized

Images