Time-frequency graph denoising method based on time-frequency matrix

Abstract

The invention belongs to the technical field of time-frequency graph denoising, and particularly relates to a time-frequency graph denoising method based on a time-frequency matrix. According to the method, the whole time-frequency matrix is arranged in an ascending order, the mean value of the first 20% of data, the maximum value of the whole time-frequency matrix, the mean value of the global maximum value and the global minimum value are taken for correction, and finally weighted averaging is carried out on the three values according to the proportion of 1: 2: 1 to obtain the dynamic threshold calculation method. According to the method, the weighted average of the global maximum value, the mean value of the first 20% of data of the time-frequency matrix and the corrected mean value serves as the interception threshold of time-frequency graph denoising, and under the low signal-to-noise ratio, the signal point detection rate of the denoising threshold of the method is higher than that of energy threshold denoising and histogram denoising. Moreover, in general, the improved denoising threshold performance provided by the method is relatively stable.

Description

technical field [0001] The invention belongs to the technical field of time-frequency image denoising, and in particular relates to a time-frequency matrix-based time-frequency image denoising method. Background technique [0002] The modern communication environment becomes more and more harsh with the development of technology, where noise is superimposed on single or multiple frequency hopping signals, which makes the extraction, sorting and parameter estimation of frequency hopping signals difficult. In order to make the parameter estimation of the frequency hopping signal more effective in the subsequent processing, it is necessary to extract a more pure frequency hopping signal in a complex environment, that is, to remove noise without causing interference and energy loss to the useful frequency hopping signal. In the context of time-frequency analysis, it is first necessary to set up a truncation threshold to denoise the obtained time-frequency diagram, so that the de...

AI Technical Summary

Problems solved by technology

[0003] Compared with the preset fixed threshold, the threshold values ​​of the energy threshold denoising algorithm and the histogram denoising algorithm are adaptive truncation thresholds that change with the change of the signal-to-noise ratio, and the performance of the two algorithms is better. But the complexity is high, and the turning point in the energy threshold denoising algorithm is not easy to locate accurately in discrete signal processing, and the distinction point between noise and signal energy distribution in the histogram denoising algorithm is also not easy to accurately locate, so this Both dynamic thresholds have certain limitations in theory and engineering practice

Images