An Infrared Image Reconstruction Method Based on Archimedes Spiral as Observation Matrix

Abstract

The invention discloses an infrared image reconstruction method based on the Archimedes spiral as an observation matrix, which includes the following steps: Step (1): read in the infrared image whose reconstruction dimension is N×N, and define it according to the dimension Archimedes spiral polar coordinate equation; step (2): uniformly collect N on the Archimedes spiral 2 points, construct a definite sequence; step (3): use the definite sequence to construct a deterministic matrix as the initial observation matrix of the infrared image whose dimension is N×N to be reconstructed; step (4): construct a definite matrix according to the sampling rate The dimension of the deterministic matrix is ​​M×N, and the deterministic observation matrix is ​​obtained, and then the infrared image is reconstructed by using the deterministic observation matrix. The invention introduces CS theory into an infrared imaging system to obtain a high-resolution reconstructed infrared image, so that the obtained reconstructed infrared image contains more valuable target information, which is helpful for subsequent target detection, identification and tracking work. And the reconstruction error is obviously reduced, and the peak signal-to-noise ratio is obviously improved.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to an infrared image reconstruction method based on an Archimedes spiral as an observation matrix. Background technique [0002] The infrared imaging system passively receives target radiation, is less affected by weather and climate conditions, has strong anti-interference ability, and has good concealment. Due to the advantages of the infrared imaging system itself, it has been widely used in civilian and military fields. Since it is difficult for traditional infrared imaging systems to reconstruct infrared images quickly and with high precision, the demand for high-resolution infrared imaging technology is imminent. [0003] In recent years, the emerging CS theory of compressed sensing obtains less observation data through projection observation, and then uses reconstruction algorithms to solve optimization problems, and reconstructs the original signal from a small a...

AI Technical Summary

Problems solved by technology

Currently Liu X J, Xia S T, Dai T's article IEEEInternational Conference on Acoustics, Speech and Signal Processing Vol. 85 No. A7, August 2015, Deterministic constructions of binary measurement matrices with various sizes proposed that although many random observation matrices have been proved to satisfy The construction conditions of the observation matrix, but some random observation matrices require a large amount of storage space, and the specific engineering implementation of the random observation matrix cannot be guaranteed

Li S, Ge G's article IEEE Transactions on Signal Processing Vol. 62 No. 11, April 2014, Deterministic Construction of Sparse Sensing Matrices via Finite Geometry proposed that although the random observation matrix satisfies the construction conditions of the observation matrix with a high probability, the current There is no feasible algorithm to check whether the random observation matrix satisfies the construction conditions of the observation matrix

Moreover, the computational complexity of the random observation matrix is ​​high, the storage space is large, and the hardware implementation is difficult

Images