Sparse Representation Analysis of Space Object Visible Imagery based on Compressive Sensing

Compressive sensing (CS) technology combines sampling and compression in the traditional information acquisition process. According to CS, the amount of measurement data is far less than that of traditional sampling, which can effectively alleviate the pressure of spaceborne data storage and process...

Full description

Saved in:
Bibliographic Details
Published in2023 8th International Conference on Intelligent Computing and Signal Processing (ICSP) pp. 1125 - 1130
Main Authors Yin, Hang, Huo, Yurong, Liu, Yuqing, Jiang, Peng
Format Conference Proceeding
LanguageEnglish
Published IEEE 21.04.2023
Subjects
compressive sensing
Image coding
Integrated optics
Memory
Optical imaging
Optical sensors
Pressure measurement
representation basis
Simulation
space object
sparse representation
Online AccessGet full text

Cover

More Information
Summary:Compressive sensing (CS) technology combines sampling and compression in the traditional information acquisition process. According to CS, the amount of measurement data is far less than that of traditional sampling, which can effectively alleviate the pressure of spaceborne data storage and processing in space observation field. Sparse representation is the prior factor of applying CS in space object visible imagery. This paper quantitatively studies the sparse representation of space object scene in various orthogonal complete bases and different representation methods. The sparse representation performance of space object image in two datasets is analyzed under different sparsity settings and evaluated by overall image error and structural similarity indexes. The simulation results show that the sparsely represented space object image can keep the image quality and structural information in most cases. The conclusion verifies the admirable sparsity of space object image for CS, and supports the further research on the space observation of compression imaging method.
DOI:10.1109/ICSP58490.2023.10248900