Instantaneous Frequency Estimation Method Of FM Signal Based On Image Processing

Among the analysis methods of non-stationary signals,the instantaneous frequency of the signal is an important characteristic parameter,and the instantaneous parameter estimation of non-stationary signals has played a key role in the development of medicine,communication and geology.During the process of signal measurement and transmission,due to improper processing methods or interference from equipment,the signal may lose some frequency components,which may reduce the accuracy of signal parameter estimation.The methods which solve signal damage problem mostly start from the signal itself,very few researchers slove the signal damage problem with image processing technology.Timefrequency analysis establishes a corresponding relationship between the signal and the timefrequency image.Signal damage can be displayed on the time-frequency image through time-frequency transformation.Through predicting and repairing the damaged timefrequency image,the instantaneous signal parameters can be estimated from the timefrequency image accurently under a certain extent.In this paper,the signal distortion area will be regarded as the decay of image information,and the image restoration algorithm will be used to predict and restore the lost information.Given that the time-frequency map has rich texture information,before using the image restoration algorithm to restore the time-frequency map,it is necessary to rely on the different size of the image color components to mark the signal in the time-frequency map as white,and process the rest of the texture information as black or gray to strengthen the signal texture direction.Experiments show that after preprocessing the time-frequency image in this paper,the effect of image repairing the time-frequency image will be better.Among the instantaneous frequency extraction methods based on time-frequency diagrams,Wigner-Hough transform is more commonly used.In the case of low signal-to-noise ratio,Wigner-Hough transform parameter extraction ability is not very good.In this paper,an interpolation algorithm is introduced to combine with the rearranged Wigner-Hough transform.Compared with directly increasing the number of sampling points of the Wigner transform,the time complexity of the method in this paper is lower.In the case of low signalto-noise ratio,the time-frequency resolution will be promoted.This article applies image processing technology to signal analysis.Based on the analysis and processing of the time-frequency diagram of the damaged signal,a new idea to solve the problem of signal damage is proposed.