Eye fundus image quality evaluation method and system, computer equipment and storage medium

Abstract

The invention discloses a fundus image quality evaluation method and system, computer equipment and a storage medium, and is applied to the technical field of medical image processing, and the method comprises the steps: controlling a focusing system to successively move to different diopter positions in a preset diopter range, and obtaining a fundus image of each diopter position; performing quality evaluation on the eye fundus image, and calculating comprehensive evaluation parameters of multiple parameters; after all preset diopter positions are traversed, determining an optimal focusing diopter position based on the comprehensive evaluation parameters, and controlling the focusing system to move to the optimal focusing diopter position; according to the method, the image quality is comprehensively evaluated through multi-parameter normalized weighted fusion, and the accuracy and efficiency of automatic focusing are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical image processing, and in particular to an eye fundus image quality evaluation method and system, computer equipment and a storage medium. BACKGROUND

[0002] An eye fundus image is an important basis for diagnosing ophthalmic diseases, and its quality directly affects the accuracy of diagnosis. In eye fundus imaging, automatic focusing technology is the key to obtaining high-quality images. Traditional automatic focusing methods usually rely on only a single indicator for evaluation, which is difficult to comprehensively reflect the image quality, resulting in inaccurate focusing or image quality that does not meet the diagnostic requirements.

[0003] To overcome these defects, the present application provides an eye fundus image quality evaluation method and system, computer equipment and a storage medium, which comprehensively evaluate the quality of eye fundus images and improve the accuracy and efficiency of automatic focusing. SUMMARY

[0004] The purpose of the present application is to provide an eye fundus image quality evaluation method and system, computer equipment and a storage medium, which aims to solve the above problems.

[0005] To achieve the above purpose, the present application provides the following technical solutions:

[0006] In a first aspect, the present application provides an eye fundus image quality evaluation method, comprising the steps of:

[0007] controlling a focusing system to move successively to different diopter positions within a preset diopter range, and obtaining eye fundus images at each diopter position;

[0008] evaluating the quality of the eye fundus images and calculating a comprehensive evaluation parameter of multiple parameters;

[0009] After traversing all the preset diopter positions, determining an optimal focusing diopter position based on the comprehensive evaluation parameter, and controlling the focusing system to move to the optimal focusing diopter position;

[0010] The quality evaluation includes region selection of the eye fundus images and normalized weighted fusion of the multiple parameters;

[0011] Specifically, the region selection of the eye fundus images includes obtaining an eye fundus image at an initial diopter position, determining a sub-region position for calculation based on the quality of the eye fundus image, controlling the focusing system to move successively to each preset diopter position, obtaining eye fundus images at different diopter positions, and cutting out images of the same sub-region position in the eye fundus images;

[0012] The normalizing and weighted fusing of the multiple parameters specifically includes that the multiple parameters at least include signal-to-noise ratio, contrast, definition and focus condition; the signal-to-noise ratio, contrast, definition and focus condition of the image of the sub-region position are respectively calculated and the multiple parameters are normalized; the normalized signal-to-noise ratio, contrast, definition and focus condition are weighted and fused according to preset weight parameters to obtain the comprehensive evaluation parameter of the fundus image quality.

[0013] In a second aspect, the application provides a fundus image quality evaluation system, specifically comprising:

[0014] a focusing module configured to control the focusing system to move within a preset diopter range;

[0015] an image acquisition module configured to acquire fundus images at different diopter positions;

[0016] an image analysis module configured to calculate the comprehensive evaluation parameter of the fundus image;

[0017] a control module configured to determine an optimal focusing diopter position based on the comprehensive evaluation parameter and control the focusing system to move to the optimal focusing diopter position.

[0018] In a third aspect, the application provides a computer device, which comprises a processor and a memory coupled to the processor, wherein the memory stores program instructions for implementing a fundus image quality evaluation method; and the processor is configured to execute the program instructions stored in the memory to implement a fundus image quality evaluation.

[0019] In a fourth aspect, the application provides a storage medium storing processor-executable program instructions for executing a fundus image quality evaluation method.

[0020] The application provides a fundus image quality evaluation method, system, computer device and storage medium, which has the following beneficial effects:

[0021] The application moves the focusing system within a preset diopter range and acquires images, and then calculates four indexes of signal-to-noise ratio, contrast, definition and focus condition after cutting the calculation region of each image, and weighted fuses the four indexes into a comprehensive score of 0-100% after normalization processing according to preset weights, and finally selects the diopter position with the highest score as the optimal focusing diopter position; the method effectively overcomes the defects of traditional single index being easily disturbed by noise, uneven light or differences in fundus structure, significantly improves the focusing accuracy and adaptability, and can be widely applied to fundus photography, OCT and other medical devices and industrial detection fields, and provides reliable technical support for high-quality imaging. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 FIG. 1 is a flowchart of an eye fundus image quality evaluation method according to an embodiment of the present application;

[0023] Figure 2 FIG. 2 is a structural diagram of an eye fundus image quality evaluation system according to an embodiment of the present application;

[0024] Figure 3 FIG. 3 is a structural diagram of a computer device according to an embodiment of the present application;

[0025] Figure 4 FIG. 4 is a structural diagram of a storage medium according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] It should be understood that the specific embodiments described herein merely exemplify the present application and are not intended to limit the present application.

[0027] The following analyzes the prior art in the related art.

[0028] The image quality evaluation method in the prior art has the following problems:

[0029] Single index limitation: only relying on a single index such as sharpness or contrast, it is difficult to comprehensively evaluate the image quality.

[0030] Noise interference: the eye fundus image is easily affected by noise, and the traditional method is difficult to effectively distinguish noise and image details.

[0031] Insufficient focusing accuracy: the traditional method cannot accurately determine the focusing position, resulting in blurred or distorted images.

[0032] Therefore, there is an urgent need for an eye fundus image quality evaluation method, system, computer device and storage medium based on multi-parameter fusion to improve the accuracy of automatic focusing and image quality.

[0033] The technical solutions in the embodiments of the present application will be described clearly and completely in the embodiments of the present application in combination with the drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0034] Embodiment 1

[0035] Please refer to Figure 1 FIG. 1 is a flowchart of an eye fundus image quality evaluation method according to an embodiment of the present application; the steps include:

[0036] S1: Control the focusing system to move to different diopter positions in a preset diopter range, and obtain fundus images at each diopter position.

[0037] System errors are eliminated by calibration, such as adjusting the initial position of the lens, eliminating gear backlash, etc., to ensure the accuracy of the focusing movement. According to the specific application scenario and requirements, the diopter range of the focusing system is determined. The preset diopter range is divided into several discrete position points, and the movement step length of the focusing system between each position point is determined. The selection of the step length needs to consider factors such as imaging quality, image acquisition speed, and system control accuracy. Too small step length will result in too long image acquisition time, and too large step length may miss the best imaging diopter position.

[0038] According to the planned diopter position point sequence, corresponding motion instructions are generated. These instructions usually include motion direction, motion distance, and motion speed, etc. The motion instructions are sent to the driving device of the focusing system, such as the motor driver. The driving device controls the motor to rotate according to the instructions, and drives the focusing component to move through the transmission mechanism, so that the focusing system moves to different preset diopter positions one by one.

[0039] During the movement of the focusing system, the actual position information of the focusing component is obtained in real time by using the position sensor, and is fed back to the control system. The control system compares the actual position with the preset diopter position, and adjusts the motion instructions through a closed-loop control algorithm to ensure that the focusing system can accurately reach each preset position. When the focusing system moves to each preset diopter position, the image acquisition device is triggered to obtain the image at the current diopter position. The timing of image acquisition needs to be accurately synchronized with the motion state of the focusing system to ensure that the obtained image is a clear image after the focusing system stabilizes. The collected images are preprocessed and stored in a designated storage device, and the diopter position information corresponding to each image is recorded.

[0040] S2: Quality evaluation is performed on the fundus images, and a comprehensive evaluation parameter of multiple parameters is calculated.

[0041] Specifically, the quality evaluation includes: region selection of the fundus image and normalized weighted fusion of the multiple parameters.

[0042] Wherein, the step of region selection of the fundus image includes: obtaining the fundus image at the initial diopter position, determining the sub-region position participating in the calculation according to the fundus image quality; control the focusing system to move to each preset diopter position one by one, and obtain fundus images at different diopter positions one by one, and cut the images of the same sub-region position in the fundus images.

[0043] The step of normalizing and weighting the multi-parameters includes: the multi-parameters at least include signal-to-noise ratio, contrast, sharpness and focus condition, the signal-to-noise ratio, the contrast, the sharpness and the focus condition of the image at the sub-region position are calculated respectively, and the multi-parameters are normalized. According to the preset weight parameter, the normalized signal-to-noise ratio, the contrast, the sharpness and the focus condition are weighted and fused to obtain the comprehensive evaluation parameter of the fundus image quality.

[0044] Further, the signal-to-noise ratio (SNR) is approximately estimated by the ratio of the variance of the signal and the noise, and the calculation formula is as follows:

[0045] SNR (dB) = 10·log 10 (σ signal 2 / σ noise 2 ),

[0046] wherein σ signal 2 is the variance of the signal, σ noise 2 is the variance of the noise, it is assumed that the signal is the information obtained by Gaussian filtering the original image, and the noise is defined as the information obtained by subtracting the signal from the original image. After the SNR is calculated, the SNR is normalized to [0, 1] by using the Sigmoid function.

[0047] The contrast (Contrast) adopts the RMS (root mean square contrast) to represent the contrast of the image, which is based on the standard deviation of the pixel intensity, and the calculation formula is as follows:

[0048]

[0049] wherein I(i,j) is the pixel value (range 0-255) of the i-th row and j-th column in the image; μ is the mean value of all pixels of the image. The theoretical maximum RMS contrast of the 8-bit image is 127.5 (when the pixel value is all 0 or 255), but the actual value is usually much smaller than this, and the actual image is generally in the range of 10-50, and the reasonable range can be calibrated by experiment or the threshold value can be set by using statistical methods such as taking the percentile of historical data when normalizing.

[0050] The sharpness (sharpness) adopts the spatial domain Tenengrad gradient method, uses the Sobel operator to calculate the gradient in the horizontal and vertical directions, and represents the sharpness by the sum of the square of the gradient amplitude. The calculation formula is as follows:

[0051]

[0052] wherein G x , G yThe horizontal and vertical gradient maps after convolution of the Sobel operator. When normalized, a reasonable range is calibrated by experiment, or a threshold is set using statistical methods (such as taking the percentile of historical data).

[0053] The focusing accuracy uses the Laplacian operator to calculate the high-frequency information of the image. The calculation formula is as follows:

[0054]

[0055] Where L(i,j) is the response map obtained after convolution of the original image and the Laplacian kernel, μ L is the mean of the Laplacian response map. When normalized, a reasonable range is calibrated by experiment, or a threshold is set using statistical methods (such as taking the percentile of historical data).

[0056] After normalizing the above indicators, the comprehensive score of 0-100% of the evaluation of the fundus image quality is obtained by weighted fusion, and the calculation formula is as follows:

[0057] QualityScore = ω1·SNR + ω2·RMS + ω3·Sharpness + ω4·VOL,

[0058] Where the weights ω1, ω2, ω3, and ω4 are 25%, 25%, 25%, and 25%, respectively.

[0059] It should be noted that the signal-to-noise ratio is an index that measures the relative relationship between the effective signal intensity and the noise intensity in the image, and its core goal is to quantify image quality: the higher the SNR, the less the noise affects the image, and the clearer the image; otherwise, the noise dominates and the image quality decreases. Contrast is a statistical index that measures the overall light-dark difference of an image, reflecting the degree of dispersion of pixel values. By calculating the deviation of pixel values from the average brightness, the dynamic range and light-dark change intensity of the image are quantified. Sharpness is an index that measures the sharpness of edges and details in an image, reflecting the high-frequency information content of the image in the spatial domain; the calculation of sharpness is usually based on spatial domain gradient analysis or frequency energy distribution, and the core idea is to capture the high-frequency information changes of the image. When focusing accurately, the light signal forms the smallest diffraction spot on the sensor, and the image high-frequency information (edge, texture) is the most abundant. By analyzing the edge intensity or high-frequency information of the image, it is determined whether the focusing position is accurate.

[0060] S3: After traversing all the preset focus positions, the optimal focus position is determined based on the comprehensive evaluation parameter, and the focusing system is controlled to move to the optimal focus position.

[0061] Specifically, the focusing system is controlled to move to a next preset diopter position, a comprehensive evaluation parameter corresponding to each preset diopter position is recorded, and steps S1-S2 are repeated until all preset diopter positions are traversed. In addition, the comprehensive evaluation parameters of all preset diopter positions are compared, and a diopter position with the highest comprehensive evaluation parameter is selected as an optimal focusing diopter position, and the focusing system is controlled to move to the optimal focusing diopter position.

[0062] In summary, in the embodiment 1 of the present application, the focusing system is moved in a preset diopter range and images are acquired, a calculation region is intercepted from each image, four indexes of signal-to-noise ratio, contrast, definition and focusing condition are comprehensively calculated, the four indexes are normalized and weighted by a preset weight to be fused into a comprehensive score of 0-100%, and finally a diopter position with the highest score is selected as an optimal focusing point. The method overcomes the defects of traditional single index being easily disturbed by noise, uneven illumination or differences in fundus structure by collaborative evaluation of multiple parameters, significantly improves focusing accuracy and adaptability, and can be widely applied to medical devices such as fundus photography and OCT and industrial detection fields, and provides reliable technical support for high-quality imaging.

[0063] Embodiment 2

[0064] Please refer to Figure 2 , which is a structural schematic diagram of an eye fundus image quality evaluation system according to the embodiment 2 of the present application; specific contents include:

[0065] a focusing module, configured to control a focusing system to move in a preset diopter range;

[0066] an image acquisition module, configured to acquire eye fundus images at different diopter positions;

[0067] an image analysis module, configured to calculate a comprehensive evaluation parameter of the eye fundus images;

[0068] a control module, configured to determine an optimal focusing diopter position based on the comprehensive evaluation parameter, and control the focusing system to move to the optimal focusing diopter position.

[0069] In the embodiment, the focusing module adopts a high-precision stepping motor or a piezoelectric ceramic to drive a focusing lens group, the movement resolution can reach ±0.01D, and the diopter range can be configured (such as -20D to +20D). By receiving an instruction of the control module, the focusing lens group is moved by a preset step at a time, and is temporarily stopped at each diopter position to ensure stable image acquisition. Zero-point calibration is performed when the system is initialized to eliminate mechanical back error.

[0070] The image acquisition module can automatically detect an effective region of the eye fundus image, and intercept a central 80% region as a calculation region. After the image quality is improved by adaptive filtering and histogram equalization, the image is transmitted to the image analysis module.

[0071] The image analysis module is used to calculate the comprehensive evaluation parameter of the image. The signal-to-noise ratio is calculated by the ratio of the variance of the signal to the variance of the noise; the RMS (Root Mean Square contrast) is used to represent the contrast of the image, which is based on the standard deviation of the pixel intensity; the spatial domain Tenengrad gradient method is used to calculate the gradient in the horizontal and vertical directions using the Sobel operator, and the sharpness is represented by the sum of the squares of the gradient amplitudes; the focusing accuracy is calculated by the Laplacian operator variance to calculate the high-frequency information of the image. After normalization, the above indicators are weighted and fused to obtain the comprehensive evaluation parameter of 0-100% of the quality of the fundus image.

[0072] The control module: after traversing all the preset diopter positions, the diopter position with the highest comprehensive evaluation parameter is selected as the optimal focusing diopter position. The high-definition fundus image after focusing is saved, and the score curve of each diopter position is displayed.

[0073] In summary, the embodiment 2 of the present application controls the focusing system to move and acquire images in the preset diopter range, and uses a multi-parameter fusion image quality evaluation method to achieve more accurate and stable automatic focusing effect. Specifically, by evaluating multiple image quality indicators, the defects of traditional single evaluation indicator being easily disturbed by noise, light and other factors are overcome, and the focusing accuracy is significantly improved; secondly, dynamic region selection and parameter normalization processing enhance the adaptability of the system to different imaging conditions and improve the focusing success rate; it can be widely used in the fields of fundus imaging, microscopic observation and other fields that require high-precision focusing.

[0074] Embodiment 3

[0075] Please refer to Figure 3 , which is a structural schematic diagram of the computer device of the embodiment 3 of the present application. The computer device 50 includes a processor 51 and a memory 52 coupled to the processor 51.

[0076] The memory 52 stores program instructions for implementing the above-mentioned fundus image quality evaluation method.

[0077] The processor 51 is configured to execute the program instructions stored in the memory 52 to implement a fundus image quality evaluation method.

[0078] The processor 51 can also be referred to as a CPU (Central Processing Unit).

[0079] The processor 51 can be an integrated circuit chip with signal processing capability. The processor 51 can also be a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. The general purpose processor can be a microprocessor or the processor can be any conventional processor.

[0080] Embodiment 4

[0081] Please refer to Figure 4 , the structure diagram of the storage medium of embodiment 4 of the application. The storage medium of the embodiment of the application stores a program file 61 capable of realizing all the methods described above, wherein the program file 61 can be stored in the storage medium in the form of a software product, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method of each embodiment of the application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes, or a computer, a server, a mobile phone, a tablet, etc.

[0082] It should be noted that in this paper, the term "include", "contain" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, device, article or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes elements inherent to such process, device, article or method. Without more limitations, the element defined by the statement "including a…" does not exclude the presence of another identical element in the process, device, article or method including the element.

[0083] The above description is only the preferred embodiment of the application, and does not limit the patent scope of the application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings, or directly or indirectly applied to other related technical fields, is also included in the patent protection scope of the application.

[0084] Although the embodiments of the application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the application, and the scope of the application is defined by the appended claims and their equivalents.

[0085] Of course, the present application also has other various embodiments, and based on the embodiments, other embodiments obtained by those skilled in the art without any creative labor are within the protection scope of the present application.

Claims

1. A method for evaluating fundus image quality, characterized in that: include: Controlling the focusing system to move to different diopter positions in a preset diopter range, and obtaining fundus images at each diopter position; Performing quality evaluation on the fundus image and calculating comprehensive evaluation parameters of multiple parameters; After traversing all preset diopter positions, determining an optimal focus diopter position based on the comprehensive evaluation parameters, and controlling the focusing system to move to the optimal focus diopter position; The quality evaluation includes: performing region selection on the fundus image and performing normalized weighted fusion on the multiple parameters; The region selection of the fundus image specifically includes: obtaining a fundus image at an initial visual position, and determining the position of a subregion involved in the calculation according to the quality of the fundus image; controlling the focusing system to sequentially traverse to each preset visual position, sequentially obtaining fundus images at different visual positions, and intercepting an image at the same subregion position in the fundus image; The normalized weighted fusion of the multiple parameters specifically includes: the multiple parameters include at least signal-to-noise ratio, contrast, clarity and focus; respectively calculating the signal-to-noise ratio, contrast, clarity and focus of the image at the sub-area position, and normalizing the multiple parameters; according to preset weight parameters, weightedly fusing the normalized signal-to-noise ratio, contrast, clarity and focus to obtain a comprehensive evaluation parameter of the fundus image quality.

2. A fundus image quality evaluation method according to claim 1, characterized in that: After traversing all preset diopter positions, the step of determining the optimal focus diopter position based on the comprehensive evaluation parameters, and controlling the focusing system to move to the optimal focus diopter position specifically includes: Controlling the focusing system to move to the next preset diopter position, and recording the comprehensive evaluation parameters corresponding to each preset diopter position until all preset diopter positions are traversed; The comprehensive evaluation parameters of all preset diopter positions are compared, the diopter position with the highest comprehensive evaluation parameter is selected as the optimal focus diopter position, and the focusing system is controlled to move to the optimal focus diopter position.

3. A fundus image quality evaluation system, characterized in that: include: A focusing module, used to control the focusing system to move within a preset visual range; An image acquisition module, used for acquiring fundus images at different visual positions; An image analysis module, used for calculating comprehensive evaluation parameters of the fundus image; A control module is used to determine an optimal focus vision position based on the comprehensive evaluation parameters, and control the focusing system to move to the optimal focus vision position.

4. A computer device, characterized in that: The computer device includes a processor and a memory coupled to the processor, wherein the memory stores program instructions for implementing a fundus image quality assessment method as described in any one of claims 1-2; and the processor is used to execute the program instructions stored in the memory to implement a fundus image quality assessment.

5. A storage medium, characterized in that: Program instructions executable by a processor are stored, and the program instructions are used to execute the fundus image quality evaluation method according to any one of claims 1 to 2.

Images