CN-120154292-B - High frame rate imaging method for improving vertical resolution of image

CN120154292BCN 120154292 BCN120154292 BCN 120154292BCN-120154292-B

Abstract

The invention relates to a high-frame-rate imaging method for improving vertical resolution of an image, which comprises the steps of collecting dynamic deformation images of a cornea section in a cornea deformation process of a patient in a process of blowing the cornea of the patient, determining a target area of the dynamic deformation images, wherein the target area is a minimum area comprising the cornea, intercepting the target area in each frame of the dynamic deformation images to form continuous frame images of the target area, carrying out high-resolution processing on the continuous frame images of the target area, and carrying out cornea biomechanical property evaluation on the continuous frame images after the high-resolution processing. The method determines the target area of the dynamic deformation image in the cornea deformation process of the patient, wherein the target area is the minimum area comprising cornea, and the data precision of the cornea area in the continuous frame image is improved by carrying out high-resolution processing on the continuous frame image of the target area, so that the accuracy of the cornea biomechanical characteristic evaluation result of the continuous frame image after the high-resolution processing is ensured.

Inventors

TIAN LEI
JIE YING
HU BI
GUO HANWEN

Assignees

北京市眼科研究所
首都医科大学附属北京同仁医院

Dates

Publication Date: 20260512
Application Date: 20250127

Claims (7)

1. A high frame rate imaging method for improving vertical resolution of an image, the method comprising: Collecting dynamic deformation images of the cornea section in the cornea deformation process of a patient in the process of blowing the cornea of the patient, wherein the dynamic deformation images are gray images; determining a target area of the dynamic deformation image, wherein the target area is a minimum area comprising a cornea; intercepting a target area in each frame of dynamic deformation image to form continuous frame images of the target area; Carrying out high resolution processing on the continuous frame images of the target area; evaluating the biomechanical characteristics of cornea of the continuous frame images after high-resolution processing; The determining the target area of the dynamic deformation image comprises the following steps: Acquiring a first frame dynamic deformation image; obtaining the marking edge of the cornea in the first frame dynamic deformation image; identifying an identification edge of the cornea in the first frame of dynamic deformation image; Forming a labeling pixel set according to the labeling edge Forming a recognition pixel set according to the recognition edge ; Differential pixel set ; According to 、 And The gray value of each element of the plurality of elements determines a change threshold; Determining a target area of the dynamic deformation image according to the change threshold; Said basis is 、 And The gray value of each element of the plurality of elements determines a change threshold value, comprising: Determination of First mean value of gray values of all elements in (B) And a first standard deviation ; Determination of A second average value of gray values of all elements in the image And a second standard deviation ; If it is Is empty set, then the change threshold is determined to be ; If it is Non-empty set, then determine Third mean value of gray values of all elements in (B) And a third standard deviation According to And (3) with The relationship between, and And Determining a change threshold; Said basis is And (3) with The relationship between, and And Determining a change threshold comprises: If it is Then determine the change threshold as ; If it is Then determine the change threshold as 。
2. The method of claim 1, wherein determining the target region of the dynamically deformed image based on the change threshold comprises: Will be The elements are classified according to columns to form a first marked pixel point set of each column ; Determining columns Minimum line number and maximum line number of all pixel points in the array; Determining the pixel point of the row with the minimum line number upwards and the pixel point of the row with the maximum line number downwards of each column as an increased pixel point, wherein the increased pixel point and the increased pixel point are Forming a second set of labeled pixels of each column ; According to columns 、 The change threshold value is used for determining change pixel points of each column; the smallest rectangle containing all columns of the changed pixels is determined as the target area.
3. The method according to claim 2, wherein the columns are arranged in accordance with 、 And the change threshold value, determining the change pixel points of each column, including: For any column, determining the column Fourth mean value of gray values of all elements in (B) ; Determining any column Fifth mean value of gray values of all elements in (3) ; If it is And is also provided with Then determine any of the columns All the pixel points in the array are the changed pixel points in any column, For a preset gray-level threshold value, ; If it is Or (b) But is provided with According to any one of the columns 、 And the change threshold value determines a change pixel point.
4. A method according to claim 3, wherein the column according to any one of the columns 、 And the change threshold value determines a change pixel point, including: Determining a set of changes ; Determination of Sixth mean value of gray values of all elements in (3) And a sixth standard deviation ; If it is Then determine any of the columns All the pixel points in the array are the changed pixel points in any column, For a predetermined degree of precision of the variation, In order to change the threshold value of the signal, Is that A first average of the gray values of all elements in (a), Is that A first standard deviation of gray values of all elements in (a); If it is Then any one of the columns is All pixel points in the array are updated to any one of the columns Repeatedly executing the determination of the pixel point with the minimum line number of each column being up to one line and the pixel point with the maximum line number being down to one line as the added pixel point, and obtaining the pixel point with the added pixel point and the pixel point with the maximum line number being down to one line Forming a second set of labeled pixels of each column Is a step and subsequent steps.
5. The method of claim 2, wherein determining a minimum rectangle containing all columns of varying pixels as the target area comprises: enlarging the variation pixel points of each column upwards and downwards respectively A pixel point, wherein, For the redundancy value to be set in advance, Based on the determination of the patient's attributes, , At the maximum value of the redundancy to be achieved, , For the minimum value of the number of pixel points between the center point of the changing pixel points of each column and the upper edge of the first frame dynamic deformation image, For the minimum value of the number of pixel points between the center point of the changing pixel points of each column and the lower edge of the first frame dynamic deformation image, A maximum value of half of the total number of the changed pixels of each column; the minimum rectangle containing all columns of enlarged changed pixels is determined as the target area.
6. The method of claim 1, wherein said high resolution processing of successive frame images of the target region comprises: For any one of the successive frame images of the target region, by the following function Performing high resolution processing; Wherein, the For frame image identification in successive frame images of the target region, For frame images Is provided with a pixel point identification in the image, For frame images Middle (f) The value of the individual pixel points is calculated, For frame images High resolution post processing item The value of the individual pixel points is calculated, In order to downsample the matrix, In order to blur the matrix, In order to penalize the parameters, Is a regular function.
7. The method of claim 6, wherein the ; Wherein, the In order to normalize the parameters, Is the coordinates of the pixel points in the image, For being based on frame images The variance estimate is obtained.

Description

High frame rate imaging method for improving vertical resolution of image Technical Field The invention relates to the technical field of image processing, in particular to a high-frame-rate imaging method for improving the vertical resolution of an image. Background Accurate measurement of corneal biomechanics is important for ophthalmic examinations and treatments, and existing corneal biomechanics are based on sequential frame images. For example, corvis ST applies an air pulse to the center of the patient's cornea while sequential frame images of the entire blow-by causing deformation of the patient's cornea are acquired by the Scheimpflug high-speed imaging technique. And analyzing the continuous frame images acquired by the image acquisition equipment to obtain parameters (such as deformation amplitude, deformation speed, curvature and the like) of biomechanics characteristics of the cornea of the patient. However, corvis ST's vertical resolution is subject to light diffraction, which results in poor resolution of the acquired image, which can affect the accuracy of analysis of the biomechanical properties of the cornea. Disclosure of Invention In order to solve the above problems, the present invention provides a high frame rate imaging method for improving vertical resolution of an image, the method comprising: collecting dynamic deformation images of the cornea section in the cornea deformation process of a patient in the process of blowing the cornea of the patient, wherein the dynamic deformation images are gray images; determining a target area of the dynamic deformation image, wherein the target area is a minimum area comprising a cornea; Intercepting a target area in each frame of dynamic deformation image to form continuous frame images of the target area; carrying out high resolution processing on continuous frame images of a target area; and evaluating the biomechanical characteristics of the cornea for the continuous frame images after the high-resolution processing. Optionally, determining the target area of the dynamic deformation image includes: Acquiring a first frame dynamic deformation image; obtaining the marking edge of the cornea in the first frame of dynamic deformation image; identifying an identification edge of the cornea in the first frame of dynamic deformation image; forming a set of labeling pixels from labeling edges Forming a set of recognition pixels from the recognition edge; Differential pixel set; According to、AndThe gray value of each element of the plurality of elements determines a change threshold; and determining a target area of the dynamic deformation image according to the change threshold value. Alternatively according to、AndThe gray value of each element of the plurality of elements determines a change threshold value, comprising: Determination of First mean value of gray values of all elements in (B)And a first standard deviation; Determination ofA second average value of gray values of all elements in the imageAnd a second standard deviation; If it isIs empty set, then the change threshold is determined to be; If it isNon-empty set, then determineThird mean value of gray values of all elements in (B)And a third standard deviationAccording toAnd (3) withThe relationship between, andAndA change threshold is determined. Alternatively according toAnd (3) withThe relationship between, andAndDetermining a change threshold comprises: If it is Then determine the change threshold as; If it isThen determine the change threshold as。 Optionally, determining the target area of the dynamic deformation image according to the change threshold value includes: Root will The elements are classified according to columns to form a first marked pixel point set of each column; Determining columnsMinimum line number and maximum line number of all pixel points in the array; Determining the pixel point of the row with the minimum line number upwards and the pixel point of the row with the maximum line number downwards of each column as an increased pixel point, wherein the increased pixel point and the increased pixel point are Forming a second set of labeled pixels of each column; According to columns、And a change threshold value, determining a change pixel point of each column; the smallest rectangle containing all columns of the changed pixels is determined as the target area. Alternatively, according to columns、And a change threshold, determining a change pixel point of each column, including: for any column, any column is determined Fourth mean value of gray values of all elements in (B); Determining any columnFifth mean value of gray values of all elements in (3); If it isAnd is also provided withThen determine any columnAll the pixel points in the array are the changed pixel points in any column,For a preset gray-level threshold value,; If it isOr (b)But is provided withAccording to any column、And determining a change pixel point by the change threshold. Optionally accordin