JP3816419B2 - Fundus image processing method, fundus image processing apparatus, recording medium thereof, and program thereof - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a fundus image processing method, a fundus image processing apparatus, a recording medium thereof, and a program thereof that can speed up image processing on the fundus image.
[0002]
[Prior art]
Conventionally, in the field of ophthalmology, the fundus is continuously photographed using a visible fluorescent agent and an infrared fluorescent agent, the appearance of the visible fluorescent agent and the infrared fluorescent agent on the fundus is observed, and lesions on the fundus are observed. Diagnosis of whether or not there is a part has been done.
[0003]
In fundus imaging using such a visible fluorescent agent and an infrared fluorescent agent, imaging is performed many times in a series of processes from when the fluorescent agent starts to appear on the fundus until the fluorescent agent disappears from the fundus.
[0004]
The luminance of the fundus image is slightly brighter in the vicinity of the nipple but slightly darker and flows to the fundus when the amount of the fluorescent agent flowing into the blood vessel of the fundus is small immediately after the fluorescent agent begins to appear in the blood vessel of the fundus As the amount of the fluorescent agent increases and the fluorescent agent spreads to the blood vessels of the fundus, the overall brightness of the fundus image increases, and after the fluorescent agent has sufficiently spread to the blood vessels of the fundus, the fluorescence that flows to the blood vessels of the fundus As the amount of the agent decreases, the overall brightness of the fundus image gradually decreases.
[0005]
When the fundus image having such a luminance change is continuously photographed, a plurality of fundus images in which the luminance of the fundus image continuously changes can be obtained.
[0006]
The plurality of fundus images are appropriately selected and used for diagnosis, but if this kind of fundus image is viewed as it is, if the luminance of the fundus image is too low, it is difficult to identify the lesion, Even if the luminance of the fundus image is too high, it is difficult to identify the lesioned part. Recently, a fundus image processing method for image processing of the luminance of the fundus image is being taken into the diagnosis of the lesioned part of the fundus.
[0007]
[Problems to be solved by the invention]
When processing the luminance of the fundus image, the shape of the nipple is relatively easy to recognize, so the outer edge of the nipple is approximated to an ellipse and the inside of the ellipse is set as a reference region. It is conceivable to perform image processing on the luminance of a lesion area, which is a specific region, using the average luminance of the image.
[0008]
By the way, in order to set the reference area, at least 3 to 4 outer edge portions, which are considered to be the outer edge of the nipple on the screen, must be designated as reference area setting coordinate points.
[0009]
In this case, one fundus image is selected from the plurality of fundus images, the nipple portion of the selected fundus image is set as the reference region, and the remaining fundus images are within the reference region for the selected image. It is conceivable to perform image processing on the luminance of a specific region for the remaining fundus image using the luminance of the image.
[0010]
However, the fundus image obtained in the field of fluorescence photography has its luminance changed with time, that is, the luminance of the nipple changes with time, so the reference for the selected fundus image If the luminance of the specific region for the remaining fundus image is image-processed using the luminance in the region, the luminance distribution of the specific region for the remaining fundus image cannot be accurately evaluated.
[0011]
Therefore, in order to accurately evaluate the luminance distribution of a specific area of a large number of fundus images, a reference area must be set for each fundus image, but a reference area is set for each fundus image. Then, it is necessary to specify the outer edge part by changing the position for each fundus image. For example, if there are nine fundus images for which the luminance distribution of a specific region must be analyzed, In order to set the reference region of the fundus image, the reference region setting coordinate points which are 36 outer edge portions must be designated in total, and the work that cannot endure the trouble is forced.
[0012]
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a fundus image processing method, a fundus image processing apparatus, and a method for speeding up image processing of the luminance distribution of the fundus image. To provide a recording medium and its processing program.
[0013]
[Means for Solving the Problems]
The image processing method according to claim 1 sets a reference area for each fundus image in order to evaluate a specific area for each of a plurality of fundus images having the same resolution, and sets a luminance value of a pixel in the reference area. An average luminance value is calculated as an average, and the luminance for each pixel in the specific region for each fundus image is divided by the average luminance of the pixels in the reference region for the fundus image to obtain a specific region for each fundus image A fundus image processing method for obtaining a luminance distribution within the fundus as a percentage of the average luminance,
A designation step for designating a selected fundus image used as a reference in setting the reference region;
In order to set a reference area for the selected fundus image, a plurality of reference area setting coordinate points specified with different positions on the screen and a remaining fundus image for setting a reference area for the remaining fundus image A reference area setting step for calculating and setting the reference area of the remaining fundus image based on at least one reference area setting coordinate point designated on the screen for
A specific area setting step for calculating and setting based on a specific area setting coordinate point specified for each fundus image and a reference area for each fundus image in order to set a specific area for each fundus image;
It is characterized by having.
[0014]
The image processing method according to claim 2, after designating at least one reference area setting coordinate point for setting a reference area for the remaining fundus image on the screen, the reference area for the selected fundus image is set. A plurality of reference area setting coordinate points for setting are specified on the screen.
[0015]
The image processing method according to claim 3, wherein a plurality of reference area setting coordinate points for setting a reference area for the selected fundus image are designated on the screen, and then a reference area is set for the remaining fundus image. At least one reference area setting coordinate point is designated on the screen.
[0016]
The image processing method according to claim 4, wherein a fundus image in which a reference area setting coordinate point for setting a reference area is not specified is sequentially displayed on the screen, and the reference area setting coordinates of each reference area are displayed. It is characterized by specifying points.
[0017]
The image processing method according to claim 5, wherein the specific area is centered on the specific area setting coordinate point based on a reference area set by calculation for each fundus image and the specific area setting coordinate point. It is characterized in that it is set by calculation as a region.
[0018]
The image processing method according to claim 6, wherein the reference region is a nipple portion of the fundus image, and the specific region is a lesion portion of the fundus image.
[0019]
The image processing method according to claim 7 is characterized in that an area of the specific region is an integral multiple of an area of the reference region.
[0020]
The image processing method according to claim 8, wherein a coordinate point in each reference region is designated as a singular point pixel, and the average luminance of the pixels in the reference region is calculated by excluding the luminance of the singular point pixel. It is characterized by seeking.
[0021]
The image processing method according to claim 9, wherein a lower limit luminance value and an upper limit luminance value are designated, and a luminance value of each pixel in a reference region that is greater than or equal to the lower limit luminance value and less than or equal to the upper limit luminance value is used, An average luminance of pixels in the reference area is obtained.
[0022]
The fundus image processing apparatus according to claim 10 sets a reference region for each fundus image and evaluates a luminance value of a pixel in the reference region in order to evaluate a specific region for each of a plurality of fundus images having the same resolution. An average luminance value is obtained by calculation as an average of each of the fundus images, and the luminance for each pixel in the specific region for each fundus image is divided by the average luminance of the pixels in the reference region for the fundus image to specify each fundus image. A fundus image processing apparatus for obtaining a luminance distribution of pixels in a region as a percentage of the average luminance,
Display means for displaying the fundus image on a screen;
A designation means for designating a reference area setting coordinate point for setting the reference area on the screen;
Reference region calculation that sets a reference region by calculation based on a plurality of reference region setting coordinate points specified at different positions in order to set a reference region for a selected fundus image used as a basis for setting a reference region Means,
A plurality of reference region settings specified by changing the positions of at least one reference region setting coordinate point specified for the remaining fundus image and the selected fundus image in order to set the reference region for the remaining fundus image Reference area calculation means for setting a reference area for the remaining fundus image based on the coordinate points for calculation,
Specific area calculation means for setting the specific area by calculation based on a specific area setting coordinate point specified for each fundus image and each reference area in order to set a specific area for each fundus image;
It is characterized by having.
[0023]
The fundus image processing apparatus according to claim 11, further comprising display control means for sequentially displaying on the screen a fundus image in which a reference area setting coordinate point for setting the reference area is not specified. .
[0024]
The fundus image processing apparatus according to claim 12, wherein a plurality of fundus images are displayed on the screen at the same time, and are displayed in association with designation of reference area setting coordinate points for each fundus image. The color of the area corresponding diagram is changed.
[0025]
The fundus image processing apparatus according to claim 13 is provided with a magnification setting unit that sets an area of the specific region to an integral multiple of an area of the reference region.
[0026]
15. The fundus image processing apparatus according to claim 14, further comprising color processing means for dividing the luminance distribution ratio for each pixel into a plurality of stages and displaying the different colors for each section on the screen. It is characterized by.
[0027]
The fundus image processing apparatus according to claim 15 is characterized in that selection of the selected fundus image is automatically executed.
[0028]
The recording medium according to claim 16 sets a reference area for each fundus image in order to evaluate a specific area for each of a plurality of fundus images having the same resolution, and averages luminance values of pixels in the reference area For each fundus image by dividing the luminance for each pixel in the specific region by the average luminance of the pixel in the reference region for the fundus image, A computer-processable recording medium on which is recorded a program for determining the luminance distribution as a percentage of the average luminance,
A designation step for designating a selected fundus image used as a reference in setting the reference region, and a plurality of reference regions designated at different positions on the screen in order to set a reference region for the selected fundus image Based on the setting coordinate point and at least one reference area setting coordinate point designated on the screen for the remaining fundus image to set the reference area for the remaining fundus image, the reference area of the remaining fundus image is determined. Calculation based on a reference area setting step to be calculated and set, a specific area setting coordinate point designated for each fundus image to set a specific area for each fundus image, and a reference area for each fundus image The specific area setting step to be set can be processed by a computer that executes the specific area setting step.
[0029]
The program according to claim 17 sets a reference area for each fundus image in order to evaluate a specific area for each of a plurality of fundus images having the same resolution, and calculates an average of luminance values of pixels in the reference area. An average luminance value is obtained by calculation, and the luminance for each pixel in the specific region for each fundus image is divided by the average luminance of the pixels in the reference region for the fundus image to calculate the average luminance value in the specific region for each fundus image. A computer-processable program for obtaining a luminance distribution as a ratio to the average luminance, the designation step for designating a selected fundus image used as a reference in setting the reference region, and a reference region for the selected fundus image In order to set a plurality of reference area setting coordinate points specified with different positions on the screen and a reference area for the remaining fundus image A reference area setting step for calculating and setting the reference area of the remaining fundus image based on at least one reference area setting coordinate point designated on the screen for the remaining fundus image, and specifying each fundus image A specific area setting step for calculating and setting based on the specific area setting coordinate points specified for each fundus image and the reference area for each fundus image in order to set the area can be processed by a computer that executes It is characterized by being.
[0030]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of a fundus image processing method, a fundus image processing apparatus, a recording medium thereof, and a program thereof according to the present invention will be described below with reference to the drawings.
[0031]
FIG. 1 shows a fundus image processing apparatus according to the present invention. In FIG. 1, 1 is a personal computer (hereinafter referred to as a personal computer) incorporating image calculation processing software, 2 is a liquid crystal display as a display device, and 3 is A keyboard 4 is a mouse.
[0032]
For example, it is assumed that nine photographed fundus images G1 to G9 are displayed on the screen 5 of the liquid crystal display 2. The fundus images G1 to G9 have the same other imaging conditions except for the time condition. Further, it is assumed that the fundus images G1 to G9 have the same resolution (horizontal resolution X, vertical resolution Y), and the same eye to be imaged is captured by the fluorescence imaging method. FIG. 2 is an enlarged view of the screen 5 of the liquid crystal display 2. In the lower right of the screen 5, recognition images G1 ′ to G9 ′ for confirming whether or not the reference area is set are arranged in correspondence with the arrangement order of the fundus images G1 to G9. The recognition images G1 ′ to G9 ′ are, for example, circles.
[0033]
First, a mouse fund 4 is used to designate a selected fundus image to be used as a reference when setting the reference region from among the nine fundus images. For this selected fundus image, it is desirable to use an image in which the nipple portion 6 as the reference region can be recognized relatively clearly and clearly. Assume that the selected fundus image is G3, for example. Reference numeral 7 denotes a blood vessel.
[0034]
Next, as shown in FIG. 3, with respect to the selected fundus image G3, coordinate points P1 (X31, Y31) to P4 (X34, Y34) is designated using mouse 4 at four locations. When the nipple portion 6 is approximated by an ellipse, at least three points are sufficient, but four points are designated here in order to reduce errors.
[0035]
For example, a location that is considered to be the outer edge of the nipple portion 6 of p1 (upper), P2 (left), P3 (lower), and P4 (right) is designated. The coordinate points P1 (X31, Y31) to P4 (X34, Y34) are used to draw an ellipse C3 as a contour locus of the reference area F3. Hereinafter, the coordinate points P1 to P4 are referred to as reference area setting coordinate points. .
[0036]
The personal computer 1 performs an operation for obtaining an ellipse C3 that is an outline of the reference region F3 based on the coordinate positions of the reference region setting coordinate points P1 (X31, Y31) to P4 (X34, Y34). For the calculation, a formula for obtaining a quadratic curve may be used. The personal computer 1 draws an ellipse C3 on the screen based on the calculation result. FIG. 4 shows only the selected fundus image G3 on which the ellipse C3 is drawn in an enlarged manner.
[0037]
Thereby, the inside of this ellipse C3 is set as the reference area F3. When obtaining the ellipse C3, the center coordinate point O3 (Ox3, Oy3) of the ellipse C3 is also obtained by computation. In this way, when the reference area F3 is set, the color of the corresponding confirmation image G3 ′ changes, and it is recognized that the reference area F3 has been set for the selected image G3.
[0038]
The luminance of each pixel inside the reference area F3 is used to calculate the average luminance value. The average luminance value may be obtained as a value obtained by dividing the sum of the luminance values of each pixel by the number of pixels, or may be obtained as a value obtained by dividing the luminance value of each pixel by the area of the reference region F.
[0039]
Next, using the mouse 4, as shown in FIG. 7, a location that is considered to be a lesion site for the selected fundus image G3, that is, a center coordinate point Q3 (Qx3, Qy3) is designated. Next, the mouse 4 is used to draw an ellipse C3 ′ that forms the outline of the specific region H3 that is a multiple of the area of the reference region F3, in correspondence with the size of the lesion site.
[0040]
Next, at least one reference area setting coordinate point is designated on the screen using the mouse 4 in order to set the reference areas F1, F2, and F4 to F9 for the remaining fundus images G1, G2, and G4 to G9. To do. Here, for example, the outer edge part on the upper side of the nipple part 6, that is, the reference region setting coordinates for the remaining fundus images G1, G2, G4 to G9 corresponding to the reference region setting coordinate point P1 for the selected fundus image G3. Each point P1 is designated.
[0041]
When the reference area setting P1 is designated for the remaining fundus images G1, G2, G4 to G9, the color of the confirmation image Gi ′ corresponding to the designated remaining fundus image Gi is sequentially displayed.
[0042]
The coordinate positions of the reference area setting coordinate points P1 for the remaining fundus images G1, G2, G4 to G9 are P1 (X11, Y11), P1 (X21, Y21), P1 (X31, Y31), P1 (X41), respectively. , Y41), P1 (X51, Y51), P1 (X61, Y61), P1 (X71, Y71), P1 (X81, Y81), P1 (X91, Y91).
[0043]
The personal computer 1 sets the reference areas F1, F2, and F4 to F9 of the remaining fundus images G1, G2, and G4 to G9 by calculation.
[0044]
That is, since the horizontal resolution X and the vertical resolution Y are the same for the fundus images G1 to G9, the microcomputer 1 has the remaining fundus images G1, G2, G4 to G9 for the horizontal coordinate position and the vertical coordinate position. Horizontal coordinate positions X11, X21, X41 to X91, vertical coordinate positions Y11, Y21, Y41 to Y91 of each reference area setting coordinate point P1 and horizontal area coordinates of the reference area setting coordinate point P1 of the selected fundus image G3 Differences D1ΔX, D1ΔY, D2ΔX, D2ΔY, D4ΔX, D4ΔY to D9ΔX, and D9ΔY from the position X31 and the vertical coordinate position Y31 are obtained. In general,
DiΔX = Xi1-X31
DiΔY = Yi1-Y31
However, i is an integer from 1 to 9, excluding 3, here.
[0045]
For example, when the reference region setting coordinate point P1 is designated for the fundus image G1, the difference is as shown in FIG.
D1ΔX = X11−X31
D1ΔY = Y11−Y31
It is.
[0046]
In FIG. 5, the blood vessel indicated by a broken line means that the luminance is low.
[0047]
Next, the personal computer 1 sets the reference area setting coordinate points P1 designated for the remaining fundus images G1, G2, G4 to G9 and the remaining reference area setting coordinate points P2 (X32, Y32) of the selected fundus image G3. , P3 (X33, Y33), P4 (X34, Y34) and the horizontal direction of the reference area setting coordinate points P2-P4 for the remaining fundus images G1, G2, G4-G9 based on the differences DiΔX, DiΔY The coordinate position and vertical coordinate position are obtained by calculation.
[0048]
For example, the calculation for setting the reference region F1 for the remaining fundus image G1 will be described.
[0049]
In this case, the reference area setting coordinate point P1 is P1 (X11, Y11) as specified.
[0050]
The remaining reference area setting coordinate points P2 (X12, Y12), P3 (X13, Y13), and P4 (X14, Y14) are obtained by the following arithmetic expression using D1ΔX and D1ΔY.
X12 = X32 + D1ΔX
Y12 = Y32 + D1ΔY
X13 = X33 + D1ΔX
Y13 = Y33 + D1ΔY
X14 = X34 + D1ΔX
Y14 = Y34 + D1ΔY
That is, if at least one reference area setting coordinate point P1 of the remaining fundus images G1, G2, G4 to G9 corresponding to the reference area setting coordinate point P1 of the selected fundus image G3 is designated, the remaining reference area setting Coordinate points P2 to P4 can be obtained by calculation.
[0051]
In general, an arithmetic expression for obtaining the reference point setting coordinate points P2 to P4 of the remaining fundus image Gi (i is an integer from 1 to 9 here) is as follows.
Xi2 = X32 + DiΔX
Yi2 = Y32 + DiΔY
Xi3 = X33 + DiΔX
Yi3 = Y33 + DiΔY
Xi4 = X34 + DiΔX
Yi4 = Y34 + DiΔY
Based on these calculation results, the personal computer 1 performs calculations for setting the reference areas F1, F2, F4 to F9, and an ellipse corresponding to the contour lines of the reference areas F1, F2, F4 to F9 on the screen. C1, C2, and C4 to C9 are drawn. Further, the center coordinate points O1, O2, and O4 to O9 of the ellipses C1, C2, and C4 to C9 are obtained by calculation.
[0052]
In order to obtain the central coordinate points O1 (Ox1, Oy1), O2 (Ox2, Oy2), O4 (Ox4, Oy4) to O9 (Ox9, Oy9), the reference area setting coordinate points P2 to P4 obtained by calculation are used. The coordinate values Xi2, Yi2, Xi3, Yi3, Xi4, Yi4 and the coordinate values Xi1, Yi1 of the designated reference area setting coordinate point P1 may be used, or the previously obtained center coordinate point O3 (Ox3 , Oy3) and the differences DiΔX and DiΔY may be used.
[0053]
At this time, the following arithmetic expression is used.
Oxi = Ox3 + DiΔX
Oyi = Oy3 + DiΔY
In FIG. 6, as an example, the relative positional relationship between the reference region setting coordinate points P2 to P4 obtained by calculation for the fundus image G1 and the reference region setting coordinate points P2 to P4 specified for the selected fundus image G3, An ellipse C1 drawn in the fundus image G1 is shown.
[0054]
Next, the personal computer 1 obtains the average luminance value of the pixels in the reference area obtained by this calculation. These results are temporarily recorded and stored in a recording medium (not shown).
[0055]
The personal computer 1 determines the center coordinate point Qi (Qxi, Qyi) of the specific regions H1, H2, H4 to H9 of the remaining fundus images G1, G2, G4 to G9, and the center coordinate point Q3 (Qx3, Qy3) and the center coordinates. It is obtained from the point O3 (Ox3, Oy3) and the central coordinate points O1 (Ox1, Oy1), O2 (Ox2, Oy2), O4 (Ox4, Oy4) to O9 (Ox9, Oy9).
[0056]
That is, the center coordinate point Qi of the specific area Hi of the remaining fundus image Gi (i is excluding 3) is
Qxi = Oxi-Ox3 + Qx3
Qyi = Oyi-Oy3 + Qy3
Next, the personal computer 1 calculates the areas S1, S2, and S4 to S9 of the reference areas F1, F2, and F4 to F9 by calculation, and the areas S1, S2, and S4 to S9 are multiplied by an integer to specify the specific areas H1, H2, The area Si of H4 to H9 is determined by calculation.
[0057]
Next, the radius of the specific area is obtained by a formula for obtaining the area of the ellipse, and an ellipse centered on the central coordinate point Qi (Qxi, Qyi) is drawn on the screen.
[0058]
Thereby, for example, as shown in FIG. 8, an ellipse C1 ′ as an outline of the specific region H1 for the fundus image G1 is drawn.
[0059]
Then, the personal computer 1 divides the luminance of each pixel in the specific areas H1, H2, and H4 to H9 by the average luminance of the pixels in the reference areas F1, F2, and F4 to F9, and obtains the division value.
[0060]
Now, assuming that the average value of the brightness of the pixels in the reference area Fi of the i-th fundus image Gi is Ri and the brightness of the pixels of the specific area Hi is R ′, the average of the reference areas Fi of the respective pixels in the specific area Hi The ratio T to luminance is
T = R '/ Ri
For each fundus image Gi, the personal computer 1 calculates a ratio T with respect to the average luminance Ri of the reference area Fi for all pixels in the specific area Hi, and displays the ratio T by coloring the pixels on the screen 5. .
[0061]
For example, as shown in FIG. 9, the luminance ratio T for each pixel in the specific area H1 is displayed in association with the color. In FIG. 9, symbol Z indicates a state in which a portion that can be identified as a lesion portion is clearly displayed.
[0062]
The association between the luminance ratio T and the color is preferably set so that a person (ophthalmologist) who performs fundus diagnosis can set as desired.
[0063]
That is, using the mouse 4, click a color processing setting icon (not shown) on the screen 5 to display a color processing setting box 10 divided into a plurality of stages on the screen 5 as shown in FIG. For example, it is desirable that the luminance distribution 0 corresponds to black, the luminance distribution 6 corresponds to white, and an appropriate color can be set therebetween.
[0064]
It is desirable that the distribution value can be set using the keyboard 3. For example, by default, the ratio T corresponds to the luminance distribution 0 (black) when the threshold value is 0.7 or less, and the ratio T is Threshold value 0.7 to threshold value 0.8 correspond to luminance distribution 1 (blue), ratio T corresponds to threshold value 0.8 to threshold value 0.9 to luminance distribution 2 (red),. .3 or more corresponds to the luminance distribution 6 (white), but in the embodiment of the present invention, these threshold values can be changed by using the keyboard 3.
(Modification 1)
As described above, in the embodiment of the invention, in order to set the reference region F3 of the selected fundus image G3, as shown in FIG. 11, the selected fundus image G3 is designated on the screen 5 (S.1), and then The reference region setting coordinate points P1 to P4 are designated by changing the position of the selected fundus image G3 (S.2), and then specific regions H1 to H9 as lesion sites for the respective fundus images G1, G2, and G4 to G9. To specify a specific region setting coordinate point Q3 (QX3, QY3) that is considered to be a lesion site for the selected fundus image G3 (S.3), and then the remaining fundus image Gi (i = 1, 2, 4 to 9), the reference area setting coordinate point P1 is designated (S.4), and then the designated reference area coordinate point P1 and the selected fundus image for the remaining fundus images G1, G2, G4 to G9. Standards specified for G3 Based on the area setting coordinate points P1 to P4, reference areas F1, F2, and F3 to F9 of the remaining fundus images G1, G2, and G4 to G9 are calculated and set (S.5), and then this specific area The specific areas H1 to H9 are set by calculation based on the setting coordinate points Q1 to Q9 and the reference areas H1 to H9 (S.6), and thereafter the luminance ratio T of each pixel in the specific area Hi is calculated. The luminance distribution in the specific area Hi is displayed on the screen 5 (S.7), but the reference area setting coordinate point P1 of the remaining fundus images G1, G2, G4 to G9 is designated first. The reference area setting coordinate points P1 to P4 of the selected fundus image G3 may be designated later.
(Modification 2)
In the embodiment of the invention, the upper and lower, right and left outer edge portions of the nipple portion 6 are designated as the reference region setting coordinate points P1 to P4, but the intersection of the blood vessel 7 and the outer edge portion 8 of the nipple 6 May be designated as the reference area setting coordinate points P1 to P4.
(Modification 3)
As shown in FIG. 12, the input box 12 for inputting the lower limit luminance value and the upper limit luminance value of the luminance value used for the calculation of the average luminance value of the pixels in the reference area F3 is displayed on the screen 5, and the keyboard 3 is displayed. The lower limit luminance value and the upper limit luminance value are input using the luminance values of the pixels within the range between the lower limit luminance value and the upper limit luminance value among the pixels in the reference area F3. An average value of the luminance of the pixels in the reference area F3 may be obtained by dividing by the number of pixels to be used.
(Modification 4)
As shown in FIG. 13, a plurality of coordinate points PQ on the blood vessel 7 that are considered to have an abnormally high luminance in the reference area Fi are designated, and pixels corresponding to the designated coordinate points PQ are designated as specific point pixels. The average luminance may be calculated by excluding the luminance value. (Modification 5)
In the embodiment of the present invention, the fundus image Gi used as a reference when setting the reference region Fi is selected by visual observation. However, for example, G4 or G5 is automatically selected from the middle of a series of images. Alternatively, it may be selected as a selected image.
[0065]
【The invention's effect】
Since the present invention is configured as described above, there is an effect that it is possible to speed up the image processing of the luminance distribution of the fundus image.
[Brief description of the drawings]
FIG. 1 is a perspective view showing a schematic configuration of a fundus image processing apparatus according to the present invention.
FIG. 2 is an enlarged view of a screen of the liquid crystal display shown in FIG.
FIG. 3 is a diagram showing reference area setting coordinate points designated on the screen of the liquid crystal display shown in FIG. 2;
FIG. 4 is an enlarged view of a selected fundus image G1 for explaining a reference region for the selected fundus image G3.
5 is an enlarged view of the fundus image G1 showing the relationship between the reference area setting coordinate points of the selected fundus image G3 shown in FIG. 3 and the reference area setting coordinate points of the fundus image G1. FIG.
FIG. 6 is an explanatory diagram for setting the reference region of the fundus image G1 by calculation using the reference region setting coordinate point of the selected fundus image G3 and the reference region setting coordinate point of the fundus image G1.
FIG. 7 is an explanatory diagram for setting a specific region of a selected fundus image G3.
FIG. 8 is an explanatory diagram for setting a specific region of a fundus image G1.
FIG. 9 is a diagram illustrating an example of a luminance distribution of a specific area obtained by calculation.
FIG. 10 is an explanatory diagram illustrating an example of setting a luminance distribution of a specific region.
FIG. 11 is a flowchart showing an example of processing steps of a fundus image processing method according to the present invention.
FIG. 12 is an explanatory diagram illustrating an example of setting a luminance distribution of a reference region.
FIG. 13 is an explanatory diagram illustrating another example of setting the luminance distribution of the reference region.
[Explanation of symbols]
1 ... Personal computer (reference area calculation means, specific area calculation means)
2. Liquid crystal display (display means)
4 ... Mouse (specifying means)
5 ... Screen
G1-G9 ... Fundus image

Claims (17)

In order to evaluate a specific area for each of a plurality of fundus images having the same resolution, a reference area is set for each fundus image, and an average luminance value is obtained by calculation as an average of the luminance values of the pixels in the reference area, For each fundus image, the luminance for each pixel in the specific region is divided by the average luminance of the pixel in the reference region for the fundus image, and the luminance distribution in the specific region for each fundus image is expressed as a percentage of the average luminance. A fundus image processing method to be obtained,
A designation step for designating a selected fundus image used as a reference in setting the reference region;
In order to set a reference area for the selected fundus image, a plurality of reference area setting coordinate points specified with different positions on the screen and a remaining fundus image for setting a reference area for the remaining fundus image A reference area setting step for calculating and setting the reference area of the remaining fundus image based on at least one reference area setting coordinate point designated on the screen for
A specific area setting step for calculating and setting based on a specific area setting coordinate point specified for each fundus image and a reference area for each fundus image in order to set a specific area for each fundus image;
A fundus image processing method comprising: After specifying at least one reference area setting coordinate point for setting a reference area for the remaining fundus image on the screen, a reference area setting coordinate point for setting a reference area for the selected fundus image is set. The fundus image processing method according to claim 1, wherein a plurality of images are designated on the screen. After a plurality of reference area setting coordinate points for setting a reference area for the selected fundus image are designated on the screen, reference area setting coordinate points for setting a reference area for the remaining fundus image are displayed on the screen. The fundus image processing method according to claim 1, wherein at least one is designated by. The fundus image in which the reference area setting coordinate point for setting the reference area is not specified is sequentially displayed on the screen, and the reference area setting coordinate point of each reference area is specified. Item 2. A fundus image processing method according to Item 1. The specific region is set by calculation based on a reference region set by calculation for each fundus image and the specific region setting coordinate point as a region centered on the specific region setting coordinate point. The fundus image processing method according to claim 1, wherein: The fundus image processing method according to claim 1, wherein the reference region is a nipple portion of the fundus image, and the specific region is a lesion site of the fundus image. The fundus image processing method according to claim 1, wherein an area of the specific region is a multiple of an area of the reference region. The coordinate point in each of the reference areas is designated as a singular point pixel, and the average luminance of the pixels in the reference area is obtained by calculation excluding the luminance of the singular point pixel. Fundus image processing method. A lower limit luminance value and an upper limit luminance value are designated, and an average luminance of pixels in the reference area is obtained using a luminance value of each pixel in the reference area that is greater than or equal to the lower limit luminance value and less than or equal to the upper limit luminance value. The fundus image processing method according to claim 1. In order to evaluate a specific area for each of a plurality of fundus images having the same resolution, a reference area is set for each fundus image, and an average luminance value is obtained by calculation as an average of the luminance values of the pixels in the reference area, For each fundus image, the luminance for each pixel in the specific region is divided by the average luminance of the pixel in the reference region for the fundus image, and the luminance distribution of the pixel in the specific region for each fundus image is relative to the average luminance. A fundus image processing apparatus to be calculated as a ratio,
Display means for displaying the fundus image on a screen;
A designation means for designating a reference area setting coordinate point for setting the reference area on the screen;
Reference region calculation that sets a reference region by calculation based on a plurality of reference region setting coordinate points specified at different positions in order to set a reference region for a selected fundus image used as a basis for setting a reference region Means,
A plurality of reference region settings specified by changing the positions of at least one reference region setting coordinate point specified for the remaining fundus image and the selected fundus image in order to set the reference region for the remaining fundus image Reference area calculation means for setting a reference area for the remaining fundus image based on the coordinate points for calculation,
Specific area calculation means for setting the specific area by calculation based on a specific area setting coordinate point specified for each fundus image and each reference area in order to set a specific area for each fundus image;
A fundus image processing apparatus. The fundus image processing apparatus according to claim 10, further comprising display control means for sequentially displaying a fundus image on which the reference area setting coordinate point for setting the reference area is not specified on the screen. A plurality of fundus images are simultaneously displayed on the screen, and the color of the reference area corresponding diagram displayed in association with the designation of the reference area setting coordinate point for each fundus image is changed. The fundus image processing apparatus according to claim 10, wherein the fundus image processing apparatus is an image processing apparatus. The fundus image processing apparatus according to claim 10, further comprising a magnification setting unit that sets an area of the specific region to an integral multiple of an area of the reference region. The fundus image according to claim 10, further comprising color processing means for dividing the luminance distribution ratio for each pixel into a plurality of stages and changing the color for each section to display on the screen. Processing equipment. The fundus image processing apparatus according to claim 10, wherein selection of the selected fundus image is automatically executed. In order to evaluate a specific area for each of a plurality of fundus images having the same resolution, a reference area is set for each fundus image, and an average luminance value is obtained by calculation as an average of the luminance values of the pixels in the reference area, For each fundus image, the luminance for each pixel in the specific region is divided by the average luminance of the pixel in the reference region for the fundus image, and the luminance distribution in the specific region for each fundus image is expressed as a percentage of the average luminance. A computer-readable recording medium on which a desired program is recorded,
A designation step for designating a selected fundus image used as a reference in setting the reference region, and a plurality of reference regions designated at different positions on the screen in order to set a reference region for the selected fundus image Based on the setting coordinate point and at least one reference area setting coordinate point designated on the screen for the remaining fundus image to set the reference area for the remaining fundus image, the reference area of the remaining fundus image is determined. Calculation based on a reference area setting step to be calculated and set, a specific area setting coordinate point designated for each fundus image to set a specific area for each fundus image, and a reference area for each fundus image And a specific area setting step to be set by the computer. In order to evaluate a specific area for each of a plurality of fundus images having the same resolution, a reference area is set for each fundus image, and an average luminance value is obtained by calculation as an average of the luminance values of the pixels in the reference area, For each fundus image, the luminance for each pixel in the specific region is divided by the average luminance of the pixel in the reference region for the fundus image, and the luminance distribution in the specific region for each fundus image is expressed as a percentage of the average luminance. A computer program that can be processed,
A designation step for designating a selected fundus image used as a reference in setting the reference region, and a plurality of reference regions designated at different positions on the screen in order to set a reference region for the selected fundus image Based on the setting coordinate point and at least one reference area setting coordinate point designated on the screen for the remaining fundus image to set the reference area for the remaining fundus image, the reference area of the remaining fundus image is determined. Calculation based on a reference area setting step to be calculated and set, a specific area setting coordinate point designated for each fundus image to set a specific area for each fundus image, and a reference area for each fundus image And a specific area setting step to be set by the computer.

Images