Research On Fast Image Processing Technology Of Swept Frequency OCT

Compared with other medical imaging technologies,Optical Coherence Tomography(OCT)has the advantages of non-invasiveness,non-contact,high resolution,and low cost.It is a medical scanning imaging technology that has developed rapidly in recent years.The imaging principle of OCT technology is to use weakly coherent light interference to illuminate the sample tissue with weak coherent light,and to detect and analyze the interference of the reflected light of different depths and the reference light to form a tomographic image.The Swept Source OCT(SS-OCT)in the OCT classification has the advantages of high imaging speed and high sensitivity.This article first compares the OCT technology with other imaging technologies,and briefly introduces the research progress of OCT system classification and related technologies at home and abroad.On the basis of analyzing the performance parameters of the sweeping OCT system,the noise model of the system is established and the speckle noise diagram of the system is simulated.According to the characteristics of the sweeping OCT system,the signal processing chain is proposed,and the efficiency of the domestic and foreign algorithms in each step is studied and compared.In order to improve the image processing speed of the sweeping OCT system,this paper proposes a fast adaptive denoising algorithm based on Wavelet Packet Decomposition and Minimum Mean Squared Filter for sweeping OCT images.Comparing the processing results of OCT image by the algorithm and Gauss filter,Wiener filter,anisotropic diffusion filter and wavelet denoising,the peak signal-to-noise ratio(PSNR)of the finger OCT image is improved by at least 22.5695 db,and the equivalent viewing number(ENL)is increased by 1.4 times.The processing time of the image is relatively short,raising the processing time of about 60 s to about 37 s.The result proves that this adaptive algorithm proposed in this paper effectively removes OCT image noise and also improves the speed of OCT image denoising.After studying the efficient image denoising algorithm,this paper also proposes an OCT image edge fast extraction algorithm based on morphological method and phase consistency principle.The algorithm first uses the morphological opening and closing algorithm to preprocess the image,uses the erosion and expansion to blur the non-edge signal information of the image,and fills the hole of the image based on the edge integrity,then uses phase consistency algorithm to extract image edges.By comparing the algorithm's edge extraction capabilities with sobel operator,roberts operator,prewitt operator,log operator,and Canny operator,it is proved that the algorithm can extract the double-edge boundary of the image well,the edge continuity is good,and the processing time is short.It is an efficient image edge extraction algorithm.