Unlock instant, AI-driven research and patent intelligence for your innovation.

GPU-based hyperspectral image RATGP and ROSP parallel optimization method

A technology of hyperspectral image and optimization method, which is applied in the field of RATGP and ROSP parallel optimization of hyperspectral image based on GPU, and can solve the time-consuming problems of serial mode

Active Publication Date: 2018-11-06
DALIAN MARITIME UNIVERSITY
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Taking RATGP (Recursive Automatic Target Generation Process, recursive unsupervised endmember extraction algorithm) and ROSP (Recursive Orthogonal Subspace Projection, parallel computing abundance estimation algorithm) algorithm as examples, the two algorithms are actually applied to endmember extraction and abundance estimation, according to the calculation principle of the algorithm, a large number of matrix operations and iterative processes are involved in the operation process, and the serial method of CPU is very time-consuming, while GPU has powerful parallel processing capability, so it is very difficult to use GPU for parallel operation. necessary

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
  • GPU-based hyperspectral image RATGP and ROSP parallel optimization method
  • GPU-based hyperspectral image RATGP and ROSP parallel optimization method
  • GPU-based hyperspectral image RATGP and ROSP parallel optimization method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the drawings in the embodiments of the present invention:

[0026] Such as figure 1 Shown: a kind of hyperspectral image RATGP and ROSP parallel optimization method based on GPU, it is characterized in that comprising:

[0027] Step 1, using GPU-based RATGP algorithm to extract endmembers. Specifically include:

[0028] S101. Set the number of endmembers num to be extracted according to the work requirements, and allocate space for the mixed pixel R on the GPU side, and use the cublasSetVector function to copy the hyperspectral mixed pixel R to the GPU memory, where R={r 1 ,r 2 ,...r i ,...,r l}, ri=(r 1 l ,r 2 l ,...,r N l ) T , l is the number of bands, N is the number of pixels, r is a vector of a mixed pixe...

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 discloses a GPU-based hyperspectral image RATGP and ROSP parallel optimization method. The method comprises the steps that a GPU-based RATGP algorithm is adopted to perform end member extraction; and a GPU-based ROSP algorithm is adopted to perform abundance estimation. According to the GPU-based RATGP algorithm, calculation is performed through a recursive form, a complicated matrixinversion calculation process is not needed, large-scale matrix calculation in the algorithm is converted into parallel calculation in a GPU, and the efficiency of the algorithm is greatly improved while precision is guaranteed. According to the GPU-based ROSP algorithm, in real-time hyperspectral de-mixing processing, an end member signal expected to be de-mixed can be enhanced step by step, theinstantaneity demand on hyperspectral remote sensing image abundance estimation is better met, and the efficiency of the algorithm is greatly improved while precision is guaranteed.

Description

technical field [0001] The invention relates to the technical field of hyperspectral remote sensing, in particular to a GPU-based hyperspectral image RATGP and ROSP parallel optimization method. Background technique [0002] Hyperspectral remote sensing is widely used in the field of earth science because of its high spatial resolution. Hyperspectral image unmixing is an important branch and research hotspot in the hyperspectral field. The hyperspectral unmixing technology includes two main steps: extraction of ground object information (ie, endmembers) and abundance estimation. However, due to the large scale of hyperspectral remote sensing data and high computational complexity, data processing efficiency has become a bottleneck restricting the further development and application of technology. [0003] In recent years, with the development of graphics processing units (graphic processing units, GPUs), large-scale data operations have been performed using the characteris...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06T1/20
CPCG06T1/20
Inventor 于纯妍宋梅萍安东张建祎王玉磊李森
Owner DALIAN MARITIME UNIVERSITY