An improved nas-rif method for blind restoration of oct images

Abstract

The invention discloses an improved NAS-RIF OCT image blind restoration method, which belongs to the field of visual detection. In the present invention, the GNLDF algorithm is used to preprocess the OCT image, and the minimum error threshold based on the Poisson distribution is used to estimate the target support domain of the OCT image; by introducing the gradient information of the OCT estimated image, an improved cost function based on ATV regularization is constructed, and the split The Bregman iterative algorithm optimizes and improves the cost function, and then adjusts the FIR filter coefficients, so that the projected image is constantly approaching the real image, and the blind restoration of the OCT image is realized. This method makes full use of the gradient information of the OCT estimated image, and effectively improves the noise robustness of the NAS-RIF method in the process of blind restoration of the OCT image by constructing an improved cost function, and can retain the detailed information of the image, and obtains the structure Clear, high-signal-to-noise ratio OCT restored images with clear edges.

Description

technical field [0001] The present invention relates to a blind restoration method of an Optical Coherence Tomography (OCT) image, which belongs to the field of visual detection, and in particular to an improved nonnegativity and support recursive inverse filtering (Nonnegativity and Support Recursive Inverse Filtering, NAS-RIF) OCT image blind restoration method. Background technique [0002] OCT technology uses the low coherence of the light source to obtain the internal structure information of the sample to be tested. It has been applied in the fields of biomedicine, clinical medicine and non-destructive testing of objects. However, the OCT system is easily affected by noise during the imaging process, resulting in image degradation, masking and blurring. The detailed information of the image directly affects the accuracy of OCT image analysis. Therefore, realizing the restoration of the OCT image and obtaining the OCT image with clear structure and clear edges and high...

AI Technical Summary

Problems solved by technology

The non-blind restoration method realizes OCT image restoration based on the Point Spread Function (PSF) of the actual OCT image. However, it is difficult to accurately obtain the PSF of the OCT image in practical applications, which makes it difficult to obtain high-quality OCT restoration using the non-blind restoration method. image; while the NAS-RIF method in the blind restoration method can use the target support domain and non-negativity of the OCT image to impose constraints on the restoration process and achieve reliable image restoration when the PSF is unknown. The process of image restoration is easily affected by the noise of the observed image, which makes the estimated image biased, and the noise robustness is low, which restricts the improvement of the quality of OCT image restoration

Images