| In addition to the signal of the image itself,the image feature is also an important amount of information for the image.However,the quality of ghost images used to extract feature signals are poor at low imaging times,and currently the image feature signal is extracted based on the image first.Therefore,it is difficult and easy to cause noise superimposed to extract the high-quality features of the signal.The wavelet transform to join the process of ghost imaging can effectively reduce the number of samples,improve the imaging quality,provide a powerful theoretical support for the study of wavelet transform and ghost imaging,and has extensive and far-reaching practical application value in military,biomedical and other fields.This paper first introduces the technology of ghost imaging and wavelet transform.Secondly,a method for directly extracting high-frequency wavelet coefficients during ghost imaging using Haar wavelet transform is proposed.The feasibility of the algorithm is verified by experiment and simulation.The relationship between the image quality and the sparsity of the object and the edge sparsity of the object is discussed.Two sets of experimental schemes were designed under visible light and low light conditions.From the two aspects of simulation and experiment,the method proposed in this paper is compared with the traditional computational ghost image method.Simulation and experimental results show that compared with the traditional method,the proposed method can significantly improve the quality of the wavelet coefficient image,breaking the limit that the original wavelet coefficients must be collected before extracting the frequency components of the wavelet coefficients,extending the field of image wavelet coefficient extraction,effectively reducing the noise introduced in the imaging process and improving the quality of the extracted image wavelet coefficients. |
