CN110037651B - Method and device for controlling quality of fundus image - Google Patents
Method and device for controlling quality of fundus image Download PDFInfo
- Publication number
- CN110037651B CN110037651B CN201810036355.4A CN201810036355A CN110037651B CN 110037651 B CN110037651 B CN 110037651B CN 201810036355 A CN201810036355 A CN 201810036355A CN 110037651 B CN110037651 B CN 110037651B
- Authority
- CN
- China
- Prior art keywords
- image
- fundus
- fundus image
- area
- quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
- A61B3/0016—Operational features thereof
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
- A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
- A61B3/14—Arrangements specially adapted for eye photography
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Biophysics (AREA)
- Ophthalmology & Optometry (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The application discloses a quality control method and a device of fundus images, wherein the method comprises the following steps: receiving a fundus image; calculating a blur value of the fundus image; extracting blood vessels from the fundus image, determining an avascular region in the fundus image, and calculating the area of the avascular region; and judging whether the image quality of the fundus image is qualified or not according to the ambiguity value and the area of the avascular region. According to the method and the device, the purpose of judging whether the fundus image is qualified or not is achieved by comparing the ambiguity value of the fundus image and the area of the avascular region, so that the technical effect of image quality control on the fundus image is achieved, and the technical problems of rapidity and stability of image quality control of the fundus image are solved.
Description
Technical Field
The application relates to the technical field of computer vision and image processing, in particular to a method and a device for controlling the quality of an eyeground image.
Background
Because the fundus image acquisition equipment does not have the function of judging the image quality, a large number of unqualified low-quality fundus images enter a later stage doctor film reading link, so that the low-quality fundus images occupy a large number of resources, the analysis efficiency of the fundus images by a doctor is reduced, and the film reading cost of the doctor on a patient is directly improved. Meanwhile, due to the fact that the judgment is affected due to poor image quality, the misjudgment risk of a doctor is increased.
The control of image quality in the related art mainly includes two ways: one is to completely depend on the experience of the operator to judge; the other is to perform judgment by analyzing a bright area or a dark area of the fundus image.
The former mode leads to the control standard of image quality unstable because of relying on artifical judgement completely, and the latter mode leads to the functioning speed slow and unstable because of the calculated amount is big, and the function judgement is single relatively moreover, is difficult to satisfy actual clinical demand.
Disclosure of Invention
The main purpose of the present application is to provide a method for controlling the quality of fundus images, so as to quickly and stably determine whether the image quality of fundus images is qualified, and avoid the low-quality images from entering the later stage of the doctor's reading process, so as to receive the cost of the doctor's reading process, and finally achieve the purpose of reducing the risk of misjudgment.
In order to achieve the above object, according to one aspect of the present application, there is provided a quality control method of a fundus image.
The quality control method of a fundus image according to the present application includes:
receiving a fundus image;
calculating a blur value of the fundus image;
extracting blood vessels from the fundus image, determining an avascular region in the fundus image, and calculating the area of the avascular region;
and judging whether the image quality of the fundus image is qualified or not according to the ambiguity value and the area of the avascular region.
Further, the determining whether the image quality of the fundus image is qualified according to the blur value and the area of the avascular region includes:
judging whether the area of the avascular region is larger than an area threshold value or not, and judging whether the ambiguity value is larger than an ambiguity threshold value or not;
and if the area of the avascular region is larger than an area threshold value and the ambiguity value is larger than the ambiguity threshold value, determining that the fundus image is a disqualified image.
Further, the method further comprises:
after the fundus image is determined to be an unqualified image, whether the blood vessel-free area comprises a light-reflecting bright area and/or a dark area is judged;
if the blood vessel-free area comprises a light-reflecting bright area and/or a dark area, the fundus image is abnormally exposed.
Further, the method further comprises:
after the fundus image is determined to be a unqualified image, whether the avascular region covers the whole fundus is also judged;
determining that the fundus image is a globally blurred image if the entire fundus is covered.
Further, the method further comprises:
after the fundus image is determined to be a global blurred image, acquiring the fundus image of the corresponding user again, and judging whether the acquired fundus image is the global blurred image or not;
and if the newly acquired fundus image is judged to be the global blurred image for a plurality of times, determining that the fundus image has fundus lesions.
Further, the determining an avascular region in the fundus image comprises:
after extracting blood vessels from the fundus image, screening the extracted blood vessels, and removing the blood vessels extracted by mistake to obtain a blood vessel area;
and determining a blood vessel-free region in the fundus image according to the obtained blood vessel region.
Further, the method further comprises:
performing color fundus image determination and anterior eye image determination on the fundus image before extracting blood vessels from the fundus image;
and if the fundus image is an achromatic fundus image or a front-eye image, rejecting the fundus image.
Further, the method further comprises:
if the fundus image is a qualified image, performing eye identification analysis on the fundus image;
performing fundus structure positioning analysis on the fundus image according to the eye classification analysis result;
judging whether the fundus image takes the fundus structure as the center according to the position of the fundus structure obtained by positioning;
and if the fundus image is not centered on the fundus structure, the fundus image determined to be qualified in image quality is finally determined to be unqualified in image quality.
In order to achieve the above object, according to another aspect of the present application, there is provided a quality control apparatus of a fundus image.
The quality control device of fundus images according to the present application includes:
an image receiving unit for receiving a fundus image;
a blur calculation unit for calculating a blur value of the fundus image;
a region extraction unit configured to extract a blood vessel from the fundus image, determine an avascular region in the fundus image, and calculate an area of the avascular region;
and the quality judgment unit is used for judging whether the image quality of the fundus image is qualified or not according to the blurring value and the area of the avascular region.
Further, the quality judgment unit includes:
the quality judgment module is used for judging whether the area of the avascular region is larger than an area threshold value and judging whether the ambiguity value is larger than an ambiguity threshold value;
and the quality determination module is used for determining the fundus image as a unqualified image if the area of the avascular region is larger than an area threshold value and the ambiguity value is larger than the ambiguity threshold value.
Further, the apparatus further comprises:
the exposure judging unit is used for judging whether the blood vessel-free area comprises a light-reflecting bright area and/or a dark area after the fundus image is determined to be an unqualified image;
an exposure determining unit for determining whether the fundus image is exposed abnormally if the avascular zone includes a light-reflecting bright zone and/or a dark zone.
Further, the apparatus further comprises:
a blur determination unit configured to determine whether the fundus image is a non-conforming image and then determine whether the avascular region covers the entire fundus;
a blur determination unit for determining the fundus image as a global blur image if the entire fundus is covered.
Further, the apparatus further comprises:
a re-judgment unit configured to re-acquire the fundus image of the corresponding user after determining that the fundus image is a globally blurred image, and judge whether the re-acquired fundus image is a globally blurred image;
a quality re-determination unit configured to determine that a fundus lesion exists in the fundus image if it is determined a plurality of times that the newly acquired fundus image is a globally blurred image.
Further, the region extraction unit includes:
the blood vessel screening module is used for screening the extracted blood vessels after the blood vessels are extracted from the fundus image, and removing the blood vessels extracted by mistake to obtain a blood vessel area;
and the avascular area module is used for determining an avascular area in the fundus image according to the obtained vascular area.
Further, the apparatus further comprises:
a color and anterior eye judgment unit configured to perform color fundus image judgment and anterior eye image judgment on the fundus image before extracting a blood vessel from the fundus image;
a chromatic and anterior eye processing unit for rejecting the fundus image if the fundus image is an achromatic fundus image or an anterior eye image.
Further, the apparatus further comprises:
the eye identification analysis unit is used for carrying out eye identification analysis on the fundus image if the fundus image is a qualified image;
the structure positioning unit is used for carrying out fundus structure positioning analysis on the fundus image according to the eye identification analysis result;
a structure determination unit configured to determine whether the fundus image is centered on the fundus structure based on the position of the fundus structure obtained by the positioning;
a quality re-determination unit for, if the fundus image is not centered on the fundus structure, finally determining a fundus image that has been determined to be of acceptable image quality as of unacceptable image quality.
In the embodiment of the application, the purpose of judging whether the fundus image is qualified is achieved by adopting the modes of calculating the ambiguity value and extracting the avascular region and comparing the ambiguity value of the fundus image with the area of the avascular region, so that the technical effect of carrying out image quality control on the fundus image is realized, and the technical problems of improving the rapidity and the stability of the image quality control of the fundus image are solved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, serve to provide a further understanding of the application and to enable other features, objects, and advantages of the application to be more apparent. The drawings and their description illustrate the embodiments of the invention and do not limit it. In the drawings:
FIG. 1 is a schematic flow chart diagram illustrating an embodiment of a method for quality control of fundus images according to the present application;
FIG. 2 is a schematic flow chart illustrating the specific working principle of one embodiment of determining whether the fundus image is acceptable;
FIG. 3 is a schematic flow chart diagram illustrating one embodiment of the present application for providing an avascular region;
FIG. 4 is a schematic flowchart of another embodiment of a method for quality control of fundus images according to the present application;
FIG. 5 is a schematic flow chart showing another embodiment of processing a fundus image using the quality control method for a fundus image;
FIG. 6 is a diagram showing the result of processing a pre-eye image screened out by yet another embodiment of the method for controlling the quality of a fundus image;
FIG. 7 is a diagram showing the results of processing a rejected image screened in accordance with still another embodiment of the method for quality control of fundus images;
FIG. 8 is a diagram showing the results of processing a qualified image screened out in yet another embodiment of a fundus image using the quality control method for fundus images;
FIG. 9 is a block diagram schematically illustrating an embodiment of the apparatus for quality control of fundus images according to the present application; and
fig. 10 is a block diagram schematically showing the structure of another embodiment of the fundus image quality control apparatus according to the present application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the terms "comprises" and "comprising," and any variations thereof, in the description and claims of this application and the above-described drawings, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
Fig. 1 is a schematic flow chart of an embodiment of a method for controlling the quality of fundus images according to the present application.
The method comprises S101-S104.
And S101, receiving a fundus image.
And S102, calculating a blurring value of the fundus image.
Specifically, a line may be extracted on the enhancement map of the fundus image in the same manner, and the gradient value of the line region may be calculated. Specifically, lines are extracted on the enhancement map by Gaussian filtering, lines are screened according to the length and the gray value of the lines, and the lines are expanded into areas; simultaneously solving a Gaussian gradient amplitude on the enhancement map, and calculating a mean value in a coverage area of the screening lines on the amplitude map; and taking logarithm of the fuzzy degree, and normalizing the logarithm of the fuzzy degree to an interval of 0-1, wherein the fuzzy degree is represented by a fuzzy factor, wherein 0 represents the most fuzzy, and 1 represents the clearest.
S103, extracting blood vessels from the fundus image, determining an avascular region in the fundus image, and calculating the area of the avascular region.
And S104, judging whether the image quality of the fundus image is qualified or not according to the blurring value and the area of the avascular region.
The captured fundus image may have problems such as blurring and exposure due to factors such as the resolution of the device for capturing the fundus image, its parameters, and the light during capturing, and therefore, it is necessary to perform image quality control on the captured fundus image to determine whether the fundus image is acceptable. The quality control method of the fundus images is applied to the field of fundus image analysis, and particularly classifies the image quality before lesion analysis is carried out on the fundus images so as to prevent unqualified fundus images from entering a lesion detection link and occupy a large amount of resources, thereby reducing the analysis efficiency of the fundus images.
The method can be applied to equipment for post-processing the shot fundus images and can also be applied to equipment for shooting the fundus images. The equipment for post-processing the fundus images comprises various equipment, such as a smart phone, a PC (personal computer), a medical analyzer and the like, the image quality of the fundus images is classified through the post-processing equipment, unqualified fundus images are removed, a detection link is carried out, and the fundus images for detecting fundus lesions are qualified high-quality images. Specifically, the fundus image may be captured by a post-processing device or may be captured by a professional imaging device such as a camera or a fundus camera. Specifically, an apparatus that performs post-processing on fundus images, such as a PC, may process one fundus image or may process a plurality of fundus images at the same time.
Fig. 2 is a schematic flow chart illustrating a specific working principle of an embodiment of determining whether the fundus image is qualified. The method comprises S201 to S202.
S201, judging whether the area of the avascular region is larger than an area threshold value or not, and judging whether the ambiguity value is larger than an ambiguity threshold value or not.
S202, if the area of the avascular region is larger than an area threshold value and the ambiguity value is larger than the ambiguity threshold value, determining that the fundus image is a disqualified image.
The utility model provides a quick classify to the eye ground image, screen qualified eye ground image to improve pathological change analysis efficiency. Therefore, it is only necessary to judge whether the image quality of the fundus image is acceptable, and it is not necessary to classify the fundus image in detail, for example, whether the fundus image is bright or dark (because it does not make much sense for actual needs). The distribution of the visible blood vessels in the fundus is very important for judging fundus lesions, so that when the area of an avascular region in the fundus image is larger than an area threshold value, the visible blood vessels in the fundus image are very few, the fundus image is very unfavorable for detecting lesions, the ambiguity directly influences the extraction of fundus structures, such as blood vessels, and when the ambiguity value is larger than the ambiguity threshold value and the area of the avascular region is larger than the area threshold value, the later-stage lesion judgment is directly influenced. Therefore, whether the image quality is qualified or not can be judged only according to the calculation of the fuzzy value and the determination of the area of the avascular region, the image quality of the fundus image can be judged quickly, and compared with manual classification, the standard is unified, so that the classification is more stable and quicker.
Up to this point, the present application has completed the determination of whether the image quality of the fundus image is acceptable by the determination of the blur value and the area of the avascular region.
In order to improve the detailed classification of the image quality of the fundus images, the reason that the fundus images are unqualified can be further judged after the judgment of whether the image quality is qualified is finished, and the reason that the fundus images are unqualified is accurately analyzed, so that the problem of the singleness of the judgment result of the fundus images in the related technology is solved. To solve this problem, the present application further analyzes the reason why the fundus image is not qualified.
The fundus image is unqualified due to various reasons, such as abnormal exposure, so in some embodiments, after the fundus image is determined to be an unqualified image, whether the non-vascular area comprises a light-reflecting bright area and/or a dark area is also judged; if the blood vessel-free area comprises a light-reflecting bright area and/or a dark area, the fundus image is abnormally exposed. The method further judges the unqualified fundus image through the embodiment, judges whether the image quality of the fundus image is unqualified or not, if the image quality is unqualified, the fundus image is subjected to exposure problem, and if the image quality is unqualified, the fundus image is adjusted when being shot through the judgment, so that the unqualified image caused by exposure is reduced.
There are various reasons for the fundus image being unacceptable, such as blurring, which directly affects the extraction of the fundus structure, so in some embodiments, after determining that the fundus image is an unacceptable image, it is also determined whether the avascular zone covers the entire fundus; determining that the fundus image is a globally blurred image if the entire fundus is covered.
The method for controlling the quality of the fundus image according to the present application may include determining exposure abnormality of an unqualified fundus image, and also include determining global blur of the unqualified fundus image; or only one part of the two is judged, namely only abnormal exposure is judged, or only global blurring is judged. This application provides the variety of judged result through the further judgement to the reason of unqualified image to the user carries out corresponding operation according to judged result, thereby has avoided the probability that unqualified eye ground image was shot in the later stage.
After the determination of the unqualified fundus image is completed through the above embodiments, since the global blur may be caused by photographing or may be caused by fundus lesions, in some embodiments, in an embodiment including the global blur determination, further, after determining that the fundus image is a global blur image, the fundus image of the corresponding user is re-acquired, and whether the re-acquired fundus image is a global blur image is determined; and if the newly acquired fundus image is judged to be the global blurred image for a plurality of times, determining that the fundus image has fundus lesions.
FIG. 3 is a flow chart illustrating one embodiment of the present application for providing an avascular region.
The method comprises S301-S302.
S301, after extracting blood vessels from the fundus image, screening the extracted blood vessels, and removing the blood vessels extracted by mistake to obtain a blood vessel area;
and S302, determining a blood vessel-free area in the fundus image according to the obtained blood vessel area.
Screening is carried out through the blood vessel of extracting in this application, gets rid of the blood vessel of mistake extraction to improve the precision that the blood vessel extracted, improve the accuracy that no blood vessel region is confirmed. Specifically, by calculating the line length, the contrast and the curvature of all the extracted blood vessels, when the line length of a certain blood vessel is smaller than a length threshold, the contrast is smaller than a contrast threshold and the curvature is smaller than a curvature threshold, the blood vessel is determined as a blood vessel extracted by mistake, the blood vessel is deleted, and the influence caused by a non-blood vessel part is eliminated, so that the accuracy of determining the area of the avascular region is improved.
The color fundus image has rich information, so the color fundus image targeted by the application is a fundus image of a non-anterior eye, and therefore before a blood vessel is extracted from the fundus image, color fundus image judgment and anterior eye image judgment are carried out on the fundus image; and if the fundus image is an achromatic fundus image or a front-eye image, rejecting the fundus image.
Specifically, the determination of whether or not the fundus image is a color fundus image may be performed by determining the number of image channels of the fundus image, determining whether or not the number of image channels is 1, and if so, determining that the fundus image is an achromatic fundus image. And the judgment before the eye can be carried out by selecting an image channel, taking the image channel B as a description, fixing a threshold (230-. If the area meeting the conditions is considered as the central reflecting area of the eyeball, the front eye is judged. It should be noted that, the judgment of the chromatic fundus image and the anterior eye image is an essential step in the present application, so as to prevent the achromatic fundus image and the non-anterior eye image from entering the detection link, and reduce the accuracy of the lesion detection in the later stage.
Fig. 4 is a flowchart illustrating another embodiment of the method for controlling quality of a fundus image according to the present application.
The method comprises S401-S404.
S401, if the fundus image is a qualified image, eye identification analysis is carried out on the fundus image, and whether the fundus image is a left eye or a right eye is analyzed. Specifically, under the condition that both the optic disc and the macular region exist, the judgment can be performed according to the positions of the optic disc and the macular region, wherein the optic disc is positioned on the left side of the macula lutea and is the left eye; the optic disc is located to the right of the macula, the right eye. Or, in the case of optic disc or macular loss, it is necessary to determine the possible position of the missing structure based on the shape of the main vascular arch and then perform the relative position determination.
And S402, performing fundus structure positioning analysis on the fundus image according to the eye identification analysis result, wherein the fundus structure specifically comprises an optic disc, a macular region, a main vascular arch and the like, which are not listed in the application.
And S403, judging whether the fundus image takes the fundus structure as the center according to the position of the fundus structure obtained by positioning.
And setting an allowable error according to the geometric position of the optic disc in the overall graph, and determining whether the image is centered on the optic disc. Or in the image positioned to the macular region, it can be determined whether the image is centered on the macula lutea only from the geometric position of the macula lutea.
And S404, if the fundus image does not take the fundus structure as the center, finally judging the fundus image which is determined to be qualified in image quality as unqualified in image quality.
The method comprises the steps of carrying out eye-to-eye analysis on qualified fundus images, carrying out eye-to-eye structure positioning according to eye-to-eye analysis results, carrying out secondary judgment on the determined qualified fundus images by judging whether the fundus images take the positioned fundus structures as the center, and improving image quality judgment precision of the fundus images through two times of image quality judgment.
FIG. 5 is a schematic flow chart showing another embodiment of processing a fundus image by applying the quality control method of a fundus image.
In the embodiment, after a group of fundus images are input, the fundus images are judged first, and non-color images and front-eye images in the group of fundus images are removed, wherein the front-eye images are shown in fig. 8.
The remaining fundus image is subjected to blur degree calculation after the non-color image and the front-eye image are removed, and simultaneously the area of the avascular region of the remaining fundus image is calculated, so that whether the image quality of the remaining fundus image is qualified or not is judged according to the blur degree value of the remaining fundus image and the area of the avascular region, and the remaining fundus image is classified into a low-quality unqualified image as shown in fig. 7 and a high-quality qualified image as shown in fig. 8.
After the fundus images are classified into unqualified images as shown in fig. 7, the method can judge the unqualified images again to judge whether the reason of the unqualified images is abnormal exposure; or whether the reason of the disqualification is the global fuzzy or not. On the basis of the determination as the global blur, it is also possible to determine again the fundus image of the global blur, thereby determining the cause of the global blur.
The fundus images judged to be qualified, such as the qualified images in the image in the.
This application carries out the reanalysis through this kind to the judged result, increases the processing mode of judged result, has not only increased the diversification of judged result, for the shooting of eye ground image provides the help, has greatly satisfied actual clinical demand to the preliminary judgement of eye ground image simultaneously.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer-executable instructions and that, although a logical order is illustrated in the flowcharts, in some cases, the steps illustrated or described may be performed in an order different than presented herein.
According to an embodiment of the present invention, there is also provided an apparatus for implementing the above-described quality control method of a fundus image, as shown in fig. 9, the apparatus including:
an image receiving unit 10 for receiving a fundus image;
a blur calculation unit 20 for calculating a blur value of the fundus image;
a region extraction unit 30 for extracting blood vessels from the fundus image, determining an avascular region in the fundus image, and calculating the area of the avascular region;
and the quality judgment unit 40 is used for judging whether the image quality of the fundus image is qualified or not according to the blurring value and the area of the avascular region.
The quality control device of the fundus image can be applied to post equipment for processing the fundus image, can also be applied to equipment for shooting the fundus image, and can control the image quality on the shooting equipment. The device for post-processing the fundus image may be a PC. Specifically, the PC is connected with the fundus camera through a network cable, after the fundus camera shoots a fundus image, the shot fundus image is sent to the PC, the PC controls the image quality of the fundus image, and whether the fundus image is a qualified image is judged.
Further, the quality judgment unit includes:
a quality determination module 401, configured to determine whether the area of the avascular region is greater than an area threshold, and determine whether the ambiguity value is greater than an ambiguity threshold;
a quality determination module 402, configured to determine that the fundus image is a failed image if the area of the avascular region is greater than an area threshold and the blurriness value is greater than the blurriness threshold.
Further, the apparatus further comprises:
an exposure judging unit 50, configured to further judge whether the avascular region includes a light-reflecting bright region and/or a dark region after determining that the fundus image is an unqualified image;
an exposure determination unit 60 for determining that the fundus image is abnormally exposed if the avascular region includes a reflective bright area and/or a dark area.
Further, the apparatus further comprises:
a blur determination unit 70 for determining whether or not the avascular region covers the entire fundus after determining that the fundus image is a non-conforming image;
a blur determination unit 80 for determining that the fundus image is a global blur image if the entire fundus is covered.
Fig. 10 is a schematic block diagram of another embodiment of the apparatus according to the present application, where the apparatus further includes:
a re-determination unit 90 configured to re-acquire the fundus image of the corresponding user after determining that the fundus image is a globally blurred image, and determine whether the re-acquired fundus image is a globally blurred image;
a quality re-determination unit 100 for determining that the fundus image has a fundus lesion if it is determined a plurality of times that the newly acquired fundus image is a globally blurred image.
Further, the region extraction unit includes:
the blood vessel screening module is used for screening the extracted blood vessels after the blood vessels are extracted from the fundus image, and removing the blood vessels extracted by mistake to obtain a blood vessel area;
and the avascular area module is used for determining an avascular area in the fundus image according to the obtained vascular area.
Further, the apparatus further comprises:
a color and anterior eye judgment unit configured to perform color fundus image judgment and anterior eye image judgment on the fundus image before extracting a blood vessel from the fundus image;
a chromatic and anterior eye processing unit for rejecting the fundus image if the fundus image is an achromatic fundus image or an anterior eye image.
Further, the apparatus further comprises:
the eye identification analysis unit is used for carrying out eye identification analysis on the fundus image if the fundus image is a qualified image;
the structure positioning unit is used for carrying out fundus structure positioning analysis on the fundus image according to the eye identification analysis result;
a structure determination unit configured to determine whether the fundus image is centered on the fundus structure based on the position of the fundus structure obtained by the positioning;
a quality re-determination unit for, if the fundus image is not centered on the fundus structure, finally determining a fundus image that has been determined to be of acceptable image quality as of unacceptable image quality. Specifically, the image receiving unit, the fuzzy calculation unit region extraction unit and the quality determination unit may be applied to an eye fundus image photographing apparatus, and the eye classification analysis unit, the structure positioning unit, the structure determination single domain and the quality re-determination unit may be applied to an apparatus for post-processing an eye fundus image; or the units can be applied to equipment for post-processing the fundus image.
The fundus image is judged twice, the first judgment is carried out according to the fuzzy value and the area of the non-vascular area of the fundus image, after the fundus image is judged to be qualified, the fundus structure is positioned according to the result of eye-based analysis, and therefore secondary judgment is carried out on whether the fundus image is qualified or not, the judgment precision of whether the fundus image is qualified or not is improved, and the effect of fundus image quality classification is improved.
From the above description, the present application can determine whether the fundus image is qualified or not by the ambiguity and the avascular area, thereby not only reducing the amount of calculation for performing qualified determination on the image, but also improving the operation speed of the system. This application not only can judge whether qualified to the eye ground image, can also judge whether unqualified reason to the user shoots again according to unqualified reason, through the judgement of this kind of unqualified reason, improves user's experience nature, can accomplish the preliminary judgement of pathological change simultaneously on unqualified basis, leads to the judgement mode of pathological change through this kind of unqualified reason, very big reduction the speed that pathological change was judged, be convenient for clinical application. Meanwhile, the adoption of twice judgment on the qualified images facilitates the feature extraction of the fundus images entering the link, and further improves the processing efficiency of the fundus images.
It will be apparent to those skilled in the art that the modules or steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present invention is not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (7)
1. A quality control method of a fundus image, characterized by comprising:
receiving a fundus image;
calculating a blur value of the fundus image;
extracting blood vessels from the fundus image, determining an avascular region in the fundus image, and calculating the area of the avascular region;
judging whether the image quality of the fundus image is qualified or not according to the ambiguity value and the area of the avascular region;
the method further comprises the following steps:
after the fundus image is determined to be a unqualified image, whether the avascular region covers the whole fundus is also judged;
if the whole fundus is covered, determining that the fundus image is a global blurred image;
after the fundus image is determined to be a global blurred image, acquiring the fundus image of the corresponding user again, and judging whether the acquired fundus image is the global blurred image or not;
if the newly acquired fundus image is judged to be a global blurred image for multiple times, determining that fundus lesions exist in the fundus image;
the method further comprises the following steps:
if the fundus image is a qualified image, performing eye identification analysis on the fundus image;
performing fundus structure positioning analysis on the fundus image according to the eye classification analysis result;
judging whether the fundus image takes the fundus structure as the center according to the position of the fundus structure obtained by positioning;
and if the fundus image is not centered on the fundus structure, the fundus image determined to be qualified in image quality is finally determined to be unqualified in image quality.
2. The method according to claim 1, wherein the determining whether the image quality of the fundus image is acceptable according to the blur value and the area of the avascular region comprises:
judging whether the area of the avascular region is larger than an area threshold value or not, and judging whether the ambiguity value is larger than an ambiguity threshold value or not;
and if the area of the avascular region is larger than an area threshold value and the ambiguity value is larger than the ambiguity threshold value, determining that the fundus image is a disqualified image.
3. The method according to claim 1 or 2, characterized in that the method further comprises:
after the fundus image is determined to be an unqualified image, whether the blood vessel-free area contains a light-reflecting bright area and/or a dark area is also judged;
if the blood vessel-free area comprises a light-reflecting bright area and/or a dark area, the fundus image is abnormally exposed.
4. The method of claim 1, wherein the determining an avascular region in the fundus image comprises:
after extracting blood vessels from the fundus image, screening the extracted blood vessels, and removing the blood vessels extracted by mistake to obtain a blood vessel area;
and determining a blood vessel-free region in the fundus image according to the obtained blood vessel region.
5. The method of claim 1, further comprising:
performing color fundus image determination and anterior eye image determination on the fundus image before extracting blood vessels from the fundus image;
and if the fundus image is an achromatic fundus image or a front-eye image, rejecting the fundus image.
6. A quality control apparatus of a fundus image, comprising:
an image receiving unit for receiving a fundus image;
a blur calculation unit for calculating a blur value of the fundus image;
a region extraction unit configured to extract a blood vessel from the fundus image, determine an avascular region in the fundus image, and calculate an area of the avascular region;
the quality judgment unit is used for judging whether the image quality of the fundus image is qualified or not according to the ambiguity value and the area of the avascular region;
the device further comprises:
a blur determination unit configured to determine whether the fundus image is a non-conforming image and then determine whether the avascular region covers the entire fundus;
a blur determination unit configured to determine that the fundus image is a global blur image if the entire fundus is covered;
a re-judgment unit configured to re-acquire the fundus image of the corresponding user after determining that the fundus image is a globally blurred image, and judge whether the re-acquired fundus image is a globally blurred image;
a quality re-determination unit configured to determine that fundus lesions exist in the fundus image if it is determined a plurality of times that the newly acquired fundus image is a globally blurred image;
the device further comprises:
the eye identification analysis unit is used for carrying out eye identification analysis on the fundus image if the fundus image is a qualified image;
the structure positioning unit is used for carrying out fundus structure positioning analysis on the fundus image according to the eye identification analysis result;
a structure determination unit configured to determine whether the fundus image is centered on the fundus structure based on the position of the fundus structure obtained by the positioning;
a quality re-determination unit for, if the fundus image is not centered on the fundus structure, finally determining a fundus image that has been determined to be of acceptable image quality as of unacceptable image quality.
7. The apparatus of claim 6, wherein the quality determination unit comprises:
the quality judgment module is used for judging whether the area of the avascular region is larger than an area threshold value and judging whether the ambiguity value is larger than an ambiguity threshold value;
and the quality determination module is used for determining the fundus image as a unqualified image if the area of the avascular region is larger than an area threshold value and the ambiguity value is larger than the ambiguity threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810036355.4A CN110037651B (en) | 2018-01-15 | 2018-01-15 | Method and device for controlling quality of fundus image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810036355.4A CN110037651B (en) | 2018-01-15 | 2018-01-15 | Method and device for controlling quality of fundus image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110037651A CN110037651A (en) | 2019-07-23 |
| CN110037651B true CN110037651B (en) | 2022-03-25 |
Family
ID=67273340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810036355.4A Active CN110037651B (en) | 2018-01-15 | 2018-01-15 | Method and device for controlling quality of fundus image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110037651B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110428410B (en) * | 2019-07-31 | 2024-02-27 | 腾讯医疗健康(深圳)有限公司 | Fundus medical image processing method, device, equipment and storage medium |
| CN110619273B (en) * | 2019-08-14 | 2023-10-31 | 张杰辉 | Efficient iris recognition method and recognition device |
| CN112508898A (en) * | 2020-11-30 | 2021-03-16 | 北京百度网讯科技有限公司 | Method and device for detecting fundus image and electronic equipment |
| CN113744254B (en) * | 2021-09-08 | 2024-06-07 | 中山大学中山眼科中心 | Fundus image analysis method, fundus image analysis system, storage medium and computer equipment |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103810492A (en) * | 2014-01-17 | 2014-05-21 | 北京大恒图像视觉有限公司 | Ambiguity analytical method of eye fundus image |
| CN105395163A (en) * | 2014-09-05 | 2016-03-16 | 佳能株式会社 | Ophthalmologic Apparatus And Ophthalmologic Apparatus Control Method |
| CN105411525A (en) * | 2015-11-10 | 2016-03-23 | 广州河谷互动医疗科技有限公司 | Eye ground picture and image intelligent obtaining and recognizing system |
| CN105787928A (en) * | 2016-02-14 | 2016-07-20 | 浙江大学 | Fuzzy fundus image automatic detection and screening method based on visual fuzziness |
| CN106650596A (en) * | 2016-10-10 | 2017-05-10 | 北京新皓然软件技术有限责任公司 | Fundus image analysis method, device and system |
| CN106651888A (en) * | 2016-09-28 | 2017-05-10 | 天津工业大学 | Color fundus image optic cup segmentation method based on multi-feature fusion |
| CN107123124A (en) * | 2017-05-04 | 2017-09-01 | 季鑫 | Retina image analysis method and device and computing equipment |
| CN107133932A (en) * | 2017-05-04 | 2017-09-05 | 季鑫 | Retina image preprocessing method and device and computing equipment |
| CN107209933A (en) * | 2014-08-25 | 2017-09-26 | 新加坡科技研究局 | For assessing retinal images and the method and system of information being obtained from retinal images |
| CN107358612A (en) * | 2017-07-07 | 2017-11-17 | 东北大学 | A kind of retinal vessel segmenting system combined based on fractal dimension with gaussian filtering and method |
| CN107451998A (en) * | 2017-08-08 | 2017-12-08 | 北京大恒普信医疗技术有限公司 | A kind of eye fundus image method of quality control |
| CN104899876B (en) * | 2015-05-18 | 2018-04-06 | 天津工业大学 | A kind of eye fundus image blood vessel segmentation method based on adaptive Gauss difference |
| CN108272434A (en) * | 2017-12-07 | 2018-07-13 | 江威 | The method and device that eye fundus image is handled |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007007044A (en) * | 2005-06-29 | 2007-01-18 | Canon Inc | Ophtalmic photographing apparatus |
| JP5702991B2 (en) * | 2010-11-19 | 2015-04-15 | キヤノン株式会社 | Image processing apparatus and image processing method |
| US9089288B2 (en) * | 2011-03-31 | 2015-07-28 | The Hong Kong Polytechnic University | Apparatus and method for non-invasive diabetic retinopathy detection and monitoring |
| CN102908120B (en) * | 2012-10-09 | 2014-09-17 | 北京大恒图像视觉有限公司 | Eye fundus image registration method, eye fundus image optic disk nerve and vessel measuring method and eye fundus image matching method |
| CN204698510U (en) * | 2015-06-02 | 2015-10-14 | 福州大学 | The diabetic retinopathy optical fundus examination photographic means that picture quality ensures |
| CN106446805B (en) * | 2016-09-08 | 2019-09-13 | 北京化工大学 | A kind of eyeground shine in optic cup dividing method and system |
| CN106530295A (en) * | 2016-11-07 | 2017-03-22 | 首都医科大学 | Fundus image classification method and device of retinopathy |
| CN107358605B (en) * | 2017-05-04 | 2018-09-21 | 深圳硅基仿生科技有限公司 | The deep neural network apparatus and system of diabetic retinopathy for identification |
-
2018
- 2018-01-15 CN CN201810036355.4A patent/CN110037651B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103810492A (en) * | 2014-01-17 | 2014-05-21 | 北京大恒图像视觉有限公司 | Ambiguity analytical method of eye fundus image |
| CN107209933A (en) * | 2014-08-25 | 2017-09-26 | 新加坡科技研究局 | For assessing retinal images and the method and system of information being obtained from retinal images |
| CN105395163A (en) * | 2014-09-05 | 2016-03-16 | 佳能株式会社 | Ophthalmologic Apparatus And Ophthalmologic Apparatus Control Method |
| CN104899876B (en) * | 2015-05-18 | 2018-04-06 | 天津工业大学 | A kind of eye fundus image blood vessel segmentation method based on adaptive Gauss difference |
| CN105411525A (en) * | 2015-11-10 | 2016-03-23 | 广州河谷互动医疗科技有限公司 | Eye ground picture and image intelligent obtaining and recognizing system |
| CN105787928A (en) * | 2016-02-14 | 2016-07-20 | 浙江大学 | Fuzzy fundus image automatic detection and screening method based on visual fuzziness |
| CN106651888A (en) * | 2016-09-28 | 2017-05-10 | 天津工业大学 | Color fundus image optic cup segmentation method based on multi-feature fusion |
| CN106650596A (en) * | 2016-10-10 | 2017-05-10 | 北京新皓然软件技术有限责任公司 | Fundus image analysis method, device and system |
| CN107123124A (en) * | 2017-05-04 | 2017-09-01 | 季鑫 | Retina image analysis method and device and computing equipment |
| CN107133932A (en) * | 2017-05-04 | 2017-09-05 | 季鑫 | Retina image preprocessing method and device and computing equipment |
| CN107358612A (en) * | 2017-07-07 | 2017-11-17 | 东北大学 | A kind of retinal vessel segmenting system combined based on fractal dimension with gaussian filtering and method |
| CN107451998A (en) * | 2017-08-08 | 2017-12-08 | 北京大恒普信医疗技术有限公司 | A kind of eye fundus image method of quality control |
| CN108272434A (en) * | 2017-12-07 | 2018-07-13 | 江威 | The method and device that eye fundus image is handled |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110037651A (en) | 2019-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107451998B (en) | Fundus image quality control method | |
| CN110037651B (en) | Method and device for controlling quality of fundus image | |
| CN110400289B (en) | Fundus image recognition method, fundus image recognition device, fundus image recognition apparatus, and fundus image recognition storage medium | |
| CN110021009B (en) | Method, device and storage medium for evaluating fundus image quality | |
| CN108346149B (en) | Image detection and processing method and device and terminal | |
| CN108272434B (en) | Method and device for processing fundus images | |
| Gagnon et al. | Procedure to detect anatomical structures in optical fundus images | |
| CN109697719B (en) | Image quality evaluation method and device and computer readable storage medium | |
| US8233674B2 (en) | Red eye false positive filtering using face location and orientation | |
| CN111368780B (en) | Combined recognition processing method and system based on fingerprint and finger vein | |
| CN108596895B (en) | Fundus image detection method, device and system based on machine learning | |
| KR102155381B1 (en) | Method, apparatus and software program for cervical cancer decision using image analysis of artificial intelligence based technology | |
| WO1999027845A1 (en) | Method of measuring the focus of close-up images of eyes | |
| CN109961848B (en) | Macular image classification method and device | |
| CN111080577A (en) | Method, system, device and storage medium for evaluating quality of fundus image | |
| CN113470102B (en) | Method, device, medium and equipment for measuring fundus blood vessel curvature with high precision | |
| CN108629378A (en) | Image-recognizing method and equipment | |
| CN106846293A (en) | Image processing method and device | |
| CN114219754A (en) | Thyroid-related eye disease identification method and device based on eye CT image | |
| CN111145155B (en) | Meibomian gland identification method | |
| CN108510446B (en) | Method and device for superimposing retinal images | |
| Ren et al. | Automatic optic disc localization and segmentation in retinal images by a line operator and level sets | |
| KR20220138069A (en) | Method, apparatus and software program for cervical cancer diagnosis using image analysis of artificial intelligence based technology | |
| CN112766193A (en) | Iris image quality evaluation method and system | |
| CN112288697A (en) | Method and device for quantifying degree of abnormality, electronic equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220412 Address after: Collective of Shangdi branch of Haidian District Talent Service Center, No. 19, Shangdi Information Road, Haidian District, Beijing 100089 Patentee after: Ke Xin Address before: 432507 group 4, Huoju village, xiaxindian Town, Yunmeng County, Xiaogan City, Hubei Province Patentee before: Jiang Wei |
|
| TR01 | Transfer of patent right |