A Compression Coding Method for Remote Sensing Signals Based on Multilevel and Fractal Vector Quantization

Abstract

The invention discloses a high-spectrum signal coding method based on multilevel and fractal dimension vector quantization. The high-spectrum signal coding method comprises the steps: firstly carrying out a de-meaning operation on an input vector of first-level vector quantization, constructing an initial code book of the first-level vector quantization, and generating a final code book and codes of the first-level vector quantization; forming a vector by a reconstructed image of the first-level vector quantization and an original de-meaning image for fractal dimension treatment, forming an initial code book of second-level vector quantization by using data of each part as an input vector of the second-level vector quantization after fractal dimension operation, and implementing quick clustering by using a quick searching algorithm to generate a final code book and indexes of the second-level vector quantization of the data of each part. According to the high-spectrum signal coding method, by combining with de-meaning, vector fractal dimension and multilevel vector quantization, the target of reducing quantization errors is achieved, thus the compression ratio and image recovering quality can be improved, the calculating complexity of the algorithm can be greatly reduced, and the purpose of carrying out quick compressed encoding on a high-spectrum remote sensing signal is achieved.

Description

technical field [0001] The invention belongs to the field of hyperspectral remote sensing image processing, in particular to a hyperspectral remote sensing image data compression encoding method combined with multi-level and fractal vector quantization techniques. Background technique [0002] All objects have the characteristics of reflecting or radiating electromagnetic waves of different wavelengths. The technology of identifying objects and their existing environment by identifying electromagnetic waves is called remote sensing. It is a comprehensive earth observation technology developed in the 1960s. It can obtain relevant data without directly contacting the target, area or phenomenon, and analyze it to obtain the required information. Remote sensing technology is based on the theory of electromagnetic radiation and integrates many disciplines such as electromagnetic wave theory, spectroscopy and colorimetry, physical and geometric optics, geography, geology, atmosphe...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the disadvantages of the existing hyperspectral remote sensing image compression using the lossless compression method, the compression ratio of the image is low and the recovery quality of the lossy compressed image is poor, and a fast compression of the hyperspectral remote sensing image based on multi-level fractal vector quantization is proposed. The encoding method improves the compression ratio and restoration quality of hyperspectral remote sensing images while keeping the size of the codebook unchanged, and also reduces the computational complexity of the algorithm

Images