Tensor decomposition cutoff remote sensing hyperspectral image compression method based on fast optimal core configuration search

A technology of tensor decomposition and image compression, applied in the field of remote sensing hyperspectral image processing

Active Publication Date: 2012-10-24
HARBIN INST OF TECH
View PDF3 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem that the current compression method based on tensor decomposition is difficult to quickly obtain the optimal tensor core configuration under the requirement

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Tensor decomposition cutoff remote sensing hyperspectral image compression method based on fast optimal core configuration search
  • Tensor decomposition cutoff remote sensing hyperspectral image compression method based on fast optimal core configuration search
  • Tensor decomposition cutoff remote sensing hyperspectral image compression method based on fast optimal core configuration search

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0068] Specific implementation mode one: the following combination figure 1 Illustrate the present embodiment, the tensor decomposition and truncation remote sensing hyperspectral image compression method based on the fast search optimal kernel configuration described in the present embodiment, its realization steps are as follows:

[0069] Step 1. Set the compression ratio C Rate and evaluation criteria;

[0070] Step 2. Take the hyperspectral image as a tensor I×J×K dimensions, I, J, and K represent space dimension 1, space dimension 2, and spectral dimension respectively;

[0071] Under the conditions of P=I, Q=J and R=K, the hyperspectral image is completely decomposed by tensor Tucker decomposition, so as to obtain the result of Tucker complete decomposition A f , B f , C f ;

[0072] Among them: P represents the space transformation dimension one of tensor decomposition, Q represents the space transformation dimension two of tensor decomposition, and R represent...

specific Embodiment approach 2

[0118] Specific implementation mode two: the following combination figure 1 Describe this embodiment, this embodiment is the set compression ratio C of Embodiment 1 Rate and further description of the evaluation criteria, the set compression ratio C described in this embodiment Rate and evaluation criteria are as follows:

[0119] Set the compression ratio C Rate The method is: normalize the elements of the nuclear tensor and the pattern matrix to 16bit; the normalization methods of the nuclear tensor and the pattern matrix are as follows:

[0120] The normalization process of the kernel tensor element G is:

[0121]

[0122] The normalization process of the mode matrix is:

[0123] u n =[u n ×10 4 ] (4)

[0124] Take the first 4 significant digits after the decimal point of each element to participate in the calculation, u n Represent pattern matrix A, B, C respectively, n=1,2,3; [] in both formulas means rounding;

[0125] After the data is normalized, it can be...

specific Embodiment approach 3

[0142] Specific implementation mode three: the following combination figure 1 Describe this embodiment. This embodiment is a further description of the method for complete decomposition using the tensor Tucker decomposition method in Embodiment 1. The method for complete decomposition using the tensor Tucker decomposition method described in this embodiment is:

[0143] Dimensionally decompose tensors into kernel tensors and orthogonal pattern matrices:

[0144]

[0145] in, is the original tensor, I×J×K dimension; is the decomposed nuclear tensor, P×Q×R dimension;

[0146] A, B, and C are orthogonal pattern matrices formed by eigenvectors arranged in columns, respectively, in I×P dimension, J×Q dimension, and K×R dimension.

[0147] The conditions for Tucker decomposition to be complete and uncompressed are: P=I, Q=J, R=K,

[0148] get

[0149] Among them: a p οb q οc r is the outer product of eigenvectors; × n Represents the modular multiplication of matrices...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a tensor decomposition cutoff remote sensing hyperspectral image compression method based on fast optimal core configuration search and relates to a hyperspectral image processing method. Aiming at the problem that a compression method based on the tensor decomposition cannot easily and fast obtain the optimal tensor core configuration under the requirement of setting the compression quality and the compression ratio, the tensor decomposition cutoff remote sensing hyperspectral image compression method based on the fast optimal core configuration search is provided. The method has the following steps that hyperspectral images are subjected to complete Tucker decomposition; spectrum dimension search starting points are calculated, iterative search is started, and the spectrum dimension optimal configuration is obtained; then, the trimming iteration is carried out, and the space dimension optimal configuration is obtained; and finally, complete decomposition results are intercepted, and final compression results are obtained. The method can be applied to satellite-bone or ground hyperspectral image compression, the compression recovery quality is ensured, and meanwhile, the calculation quantity of the compression method can be effectively reduced.

Description

technical field [0001] The invention relates to a remote sensing hyperspectral image processing method, which belongs to the field of image processing. Background technique [0002] Hyperspectral images are currently one of the research hotspots in the field of remote sensing at home and abroad. Its outstanding feature is the combination of imaging technology and spectral technology, which realizes the simultaneous acquisition of spatial information and spectral information of ground objects, that is, the integration of maps and spectra. Has been widely used in environmental monitoring, geology, meteorology, military reconnaissance and other fields. However, the superiority of hyperspectral images comes at the cost of larger data volume and higher data dimensions. For example, a standard hyperspectral AVIRIS (Airborne Visible / Infrared Imaging Spectrometer) image has 224 consecutive spectral segments, and the image spatial resolution of each spectral segment is 512×614×16bit...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06T9/00
Inventor 陈浩王佳斌周爽张晔
Owner HARBIN INST OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap