JP2001137191A - Ophthalmic imaging equipment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a good quality and stable image in a short time for an optometric image necessary for essential diagnosis. SOLUTION: Pushing the imaging switch starts imaging operation and all of the detected image information is input into image memory as digital data. The image processing section detects the highest and the lowest luminance values for the preset area of interest. The signal compression ratio of the digital signal is set based on these highest and lowest values to obtain an image with a higher contrast compared with an image necessary for essential diagnosis.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an ophthalmologic photographing apparatus having an image pickup device capable of picking up an image of a subject's eye.

[0002]

2. Description of the Related Art There is known an ophthalmologic photographing apparatus which obtains an image of an eye to be examined as an electronic image by using an image sensor or the like instead of a conventional silver halide film.

[0003]

However, in the above conventional example, when the image of the eye to be inspected is captured as an electronic image, the latitude becomes narrower than that of the conventional silver halide film. It is very difficult to efficiently obtain information from a high-brightness nipple containing a lot of information and a low-brightness retina, particularly a macula. Furthermore,
The reflectance of the fundus varies greatly from person to person, and if an image is taken with an inappropriate amount of light for this reflectance, an inefficient image with a small area containing effective information for the prepared dynamic range is obtained. Become.

Further, when an eye image to be inspected is taken, white flare may occur around the image due to the size of the curvature of the cornea or an erroneous photographing. In such a case, an unnecessary high dynamic range is required in advance. Since the information of the luminance portion is faithfully allocated, an inefficient image is obtained. Similarly, a black aperture portion at the periphery of the image also causes an inefficient image in which unnecessary information is allocated to the low-luminance portion.

[0005] Further, in general electronic images, if the image data amount is large, a large storage space is required, and the subsequent processing such as reproduction is slowed down. Also, depending on the ophthalmologic imaging device, there is a method of continuously taking images at very short intervals,
In that case, there arises a problem that the processing speed per image must be greatly reduced.

SUMMARY OF THE INVENTION An object of the present invention is to provide an ophthalmologic photographing apparatus capable of shortening the image processing time and minimizing the deterioration of the image quality of the originally required eye image area in the image. .

[0007]

An ophthalmologic photographing apparatus according to the present invention for achieving the above object has an image pickup device having a plurality of pixels for converting an eye image to be examined into an electric signal; A digital conversion unit that converts an output into a digital signal; and a first and a second signal that respectively detect a high level signal and a low level signal among level signals indicating a luminance level of the digital signal in a predetermined area pixel of the image sensor. 2
And setting means for setting the signal compression ratio of the digital signal of the digital conversion unit based on the high level signal and the low level signal.

[0008] In particular, the setting means may include a compression ratio of a level signal by the digital converter between the high level signal and the low level signal, and a compression ratio not between the high level signal and the low level signal. The compression ratio of the level signal by the digital conversion unit is different.

[0009] The setting means may set a compression ratio of a level signal by the digital conversion section between the high level signal and the low level signal to a digital level which is not between the high level signal and the low level signal. It is set smaller than the compression ratio of the level signal by the converter.

The predetermined pixel area for detecting the high level signal includes a nipple of the eye to be inspected, and the predetermined pixel area for detecting the low level signal includes a macula of the eye to be inspected. I have.

In addition, the apparatus has a fixation target for guiding a fixation position of the eye to be inspected, and the predetermined pixel area is changed according to the fixation table. On the other hand, the predetermined pixel area is a predetermined fixed area.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail with reference to the illustrated embodiment.

FIG. 1 shows a configuration diagram of a fundus camera according to a first embodiment. A condenser lens 3 is provided on an optical path from an observation light source (infrared light source) 1 to an objective lens 2 facing an eye E to be examined. A photographing light source (visible light source) 4, a mirror 5, a relay lens 6, and a perforated mirror 7 are sequentially arranged. On the optical path behind the perforated mirror 7, a focus lens 8, a photographing lens 9, and a half A mirror 10 and an imaging unit 11 are arranged, and a fixation target 12 is arranged in the incident direction of the half mirror 10. Note that the imaging device is arranged conjugate with the fundus.

Reference numeral 11 denotes an image pickup means for color photographing in which pixel rows are two-dimensionally arranged.
A (analog) / D each having a resolution of, for example, 10 bits
The digital signal is converted into a digital signal (level signal) by the (digital) converter 13. A (analog) / D (digital) converter 1
3 is connected to the image memory 14 and the image processing unit 15, and the outputs of the image memory 14 and the image processing unit 15
6 is connected to a television monitor 17. Also,
The output of the image processing unit 15 is connected to a shooting operation control unit 18 and a region of interest setting unit 19, the output of the shooting switch 20 is connected to the shooting operation control unit 18, and the output of the shooting operation control unit 18 is a light source for shooting. 4 is connected. Reference numeral 21 denotes a detection switch for detecting whether the target fundus is the left eye or the right eye.

The near-infrared light beam emitted from the observation light source 1 passes through a condenser lens 3, a mirror 5, and a relay lens 6, and
The light is reflected around the perforated mirror 7 and illuminates the fundus Er through the objective lens 2 and the pupil Ep of the eye E to be examined. The illuminated fundus image passes through the pupil Ep, the objective lens 2, and the hole of the perforated mirror 7, passes through the focus lens 8, the photographing lens 9, and the half mirror 10, and forms an image on the imaging unit 11. In addition, the subject can look at the fixation target 12 by the optical path reflecting the half mirror 10, whereby the line of sight is guided and the position of the eye E is fixed.

The image formed on the image pickup means 11 is continuously displayed on a television monitor 17 through an A / D conversion section 13, an image memory 14, and a D / A conversion means 16, and the operator displays this image. Positioning is performed while watching, and fundus photography is performed. In addition, when performing positioning, observation is performed in monochrome using near-infrared light.

When the photographing switch 20 is depressed at the time of completion of the alignment, the photographing operation control means 18 detects the depression and the photographing light source 4 emits light by a preset light emission energy. The photographing light follows the same optical path as the observation light after the mirror 5, and forms a fundus image of the eye E on the imaging unit 11. The obtained video signal (R, G, B signal) is converted into digital information by the A / D converter 13 at a gradation of, for example, 10 bits for each pixel, and is temporarily recorded in the image memory 14.
The image processing unit 15 starts the image processing.

FIG. 2 is a flowchart of the image processing. First, when the photographing switch 20 is depressed, a photographing operation is started (step 1), and when this is detected, all the accumulated pixel signals are 10 bits. All the image information is input to the image memory 14 after being converted into digital data by the gradation (step 2). Next, the left and right eye detection means 21 detects the position of the optical system with respect to the eye E to be examined, and discriminates between the left and right eyes (step 3). This left and right eye detecting means 2
The luminance information is detected for a preset region of interest in accordance with the output of (1).

FIGS. 3 and 4 show fundus images on the image sensor obtained by presenting the fixation target 12 and centering the image on the back pole of the fundus Er. The area is set to P and Q. FIG. 3 shows an image when the left eye is photographed, and FIG. 4 shows an image when the right eye is photographed. These images are for the right eye,
Two light sources of different colors may be prepared for the left eye, and the operator may prompt the subject to instruct which fixation target 12 to look at, or the operator may select the right and left to select the right or left. The two light sources may be switched according to this selection.

As shown in FIG. 3, the macula (x) is included in the area Q, the nipple (o) is included in the area P, and these areas P and Q are set as the regions of interest. However, with respect to the set regions of interest (areas Q and P), the luminance levels of all or some portions in each region are detected. It is known that the brightness of the normal nipple is high and the brightness of the macula is low. Therefore, when the luminance in the area P is higher than the luminance in the area Q, it is possible to determine that the subject's eye is the left eye, and it is possible to omit the above-described left and right eye detection switches. Then, the highest luminance signal N is detected from the plurality of pixel signals in the area P, and the lowest luminance M is detected from the plurality of pixel signals in the area Q (step 4).
・ 5).

Conversely, when the fundus of the eye to be examined is the right eye, the macula is in the P region and the nipple is in the Q region.
In the area Q, the maximum luminance signal N
Will be detected.

FIG. 5 is a histogram diagram of the luminance of the entire fundus image (corresponding to the entire image pickup device 11), wherein the horizontal axis represents luminance and the vertical axis represents frequency. Each mountain represents an aperture portion A, a retina portion B, a nipple C, and a flare D generated due to a photographing error or the like from the side of low luminance. The area indicated by the dotted line indicates luminance information detected from the areas P and Q for the left eye. As described above, in the present embodiment, the position where the macula and the nipple are present is determined based on the digital data of only this region from a predetermined relationship on the imaging surface without performing the detection operation on the entire image. The luminance part M and the maximum luminance part N are detected. Also, a useful dynamic range H can be calculated from the lowest luminance portion M and the highest luminance portion N (step 6). The compression ratio is high for the areas G and I, and the compression ratio is high for the area H. The dynamic range compression is performed on, for example, 8-bit data whose rate is set low (steps 7.8). Specifically, a conversion table as shown in FIG. 6A is prepared by the image processing unit 15. M and N in FIG. 6 correspond to M and N in FIG. 4, respectively. After performing the compression process using this conversion table, the image data is stored again in the image memory 14 (step 9).

More specifically, in this embodiment, the regions P and Q
When the highest luminance value N and the lowest luminance value M are detected, the compression ratio is reduced for digital data between the highest luminance value N and the lowest luminance value M of all pixel outputs of the image sensor, and Digital data that is not between the luminance value N and the minimum luminance value M is subjected to processing for increasing the compression ratio. This is to obtain an image having a fine gradation with respect to the digital data of the fundus image that is originally required for diagnosis and is between the highest luminance value N and the lowest luminance value M.

On the other hand, apertures and flares are unnecessary image information for the examiner, and signal processing of rough gradation is performed on these images.

In this way, it is possible to perform efficient detection in a short period of time without investigating all image information, and to generate effective information with a small amount of data. In the dynamic range compression method, the compression ratio is set higher than that of the areas G and I. However, as shown in FIG. 6B, the information in the area G is shifted to the lowest luminance value, and the information in the area I is shifted. The compression may be performed by shifting the information to the highest luminance value. The information in the area H may not be compressed.

In this embodiment, the A / D converter 13 quantizes the data at a resolution of 10 bits (higher bits) and converts it into 8-bit data (lower bits) by the image processor 15. The compression processing may be performed on the areas G and I, and the information in the area H may be decompressed. In addition, when the region of interest is first set, the lowest luminance is detected from the area Q including the macula, but when the luminance level is lower in the peripheral part of the image due to the aberration of the optical system of the fundus Er, etc. May set the area.

It should be noted that the fundus image subjected to such image processing is
The image is stored in the image memory 14. The image may be displayed on the television monitor 17, printed out by a printer (not shown), or a portable recording medium such as a magneto-optical disk (not shown). Or transfer to an external device using a communication unit. In the present embodiment, the lowest luminance part detected from the areas P and Q is detected as M, and the highest luminance part is detected as N. However, the area may be further divided into smaller parts and the average value of each part may be compared and examined. Then, set the area in advance to be small and calculate M,
N may be obtained.

For example, if detection is performed from the average value of upper or lower several pixels, or if extraneous data is excluded, stable detection of M and N without being affected by noise components becomes possible.

FIG. 7 shows a histogram when the obtained image is too dark due to a mistake in setting the amount of photographing light or individual differences in the fundus Er. As in FIG. B, a nipple C, a flare D caused by a shooting error, and the like. In this case as well, similarly, the highest luminance portion N and the lowest luminance portion M in the areas P and Q are detected, and as shown in FIG.
By creating a conversion table so that the compression ratio becomes low in the range H up to, the image can be efficiently created without being influenced by a mistake in setting the amount of photographing light or individual differences in the fundus Er. .

On the other hand, on the contrary, an incorrect setting of the photographing light amount or the fundus E
Even when an image that is too clear is obtained due to the individual difference of r, it can be similarly corrected. In addition, although the nipple and the macula or the surrounding area were set as the region of interest,
The operator may prepare a region-of-interest changing means for setting a region of particular interest, and calculate and obtain the region based on the output of the left and right eye detecting units. For example, when the operator particularly wants to pay attention to the nipple, when the left eye is detected, the P region in FIG. 3 can be determined to be the target portion, and the compression ratio of the peak of the nipple C in FIG. Can be achieved.

FIG. 9 is a block diagram showing the configuration of the second embodiment.
In this embodiment, the fixation target 12 is movable, and a fixation position detection unit 22 is added. The output of the fixation position detection unit 22 is output from the image processing unit 15 and the fixation target 1.
2 driving means 12 '.

As in the first embodiment, the subject has the fixation target 1
2, the fixation target 12 can be freely changed in position, so that an image as shown in FIG. 10, for example, corresponding to the fixation target position can be obtained. The image processing unit 15 that has detected that such an image has been obtained is the fixation position detection unit 2
From the output of No. 2, it is calculated that the macula is located substantially at the center of the area U shown in FIG. 10 in consideration of the magnification and the moving distance of the optical system. Based on the position of the macula, the nipple is expected to be in the area V or V ′. In this case, since V ′ is outside the image area, the nipple position can be determined as the area V. Note that, depending on the position of the fixation target 12, detection may be performed for both the area V or V ′, and if the left and right eye detection means 21 is provided as in the second embodiment, the nipple position is obtained from both information. Can be determined.

Thereafter, as in the first embodiment, the maximum luminance and the minimum luminance are calculated from the set region of interest, and the dynamic range is obtained to perform the dynamic range compression. By such a procedure, even in a device in which the fixation target can be freely moved, efficient detection can be performed in a short time without detecting all image information, and effective information can be obtained from a small amount of data. Can be generated.

Although the nipple and the macula are set as the regions of interest, a region of interest changing means for setting a region that the operator particularly pays attention to is prepared, and the region is calculated from the output of the fixation position detection unit 22. You may ask.

FIG. 11 shows a region of interest according to the third embodiment.
An example is shown in which a region of interest W is set for a central elongated part including the macula and the nipple in the obtained fundus images. The signal processing unit 15 obtains luminance information in the area W and creates a histogram. FIG. 12 shows a histogram created for all the regions. By detecting the inside of the region W, the approximate shapes of the peaks of the retinal part B and the nipple C can be extracted. Is divided into areas J, K, and L based on a preset reference value.

FIG. 13 is a conversion table in which the compression ratio is set low for an area J corresponding to a macula and an area L corresponding to a nipple, which contain a lot of information. In this way, the information of the fundus Er is divided into almost three areas, and different compression ratios are set to perform processing on the eye image information, so that a notable part is set in more detail, and the eye E of the eye E is examined. Necessary information can be efficiently and effectively generated without being influenced by individual differences or setting mistakes of the photographing light amount.

[0037]

As described above, in the ophthalmologic photographing apparatus according to the present invention, the signal compression ratio of the digital signal of the image sensor is set based on the high level signal and the low level signal of the predetermined area in all the pixels, especially The compression ratio for the level signal between the high level signal and the low level signal is low, and the compression ratio for the level signal not between the high level signal and the low level signal is high, which is necessary for diagnosis. It is possible to obtain an image with a high gradation for the image.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a first embodiment.

FIG. 2 is a flowchart of image processing.

FIG. 3 is an explanatory diagram of a fundus image.

FIG. 4 is an explanatory diagram of a fundus image.

FIG. 5 is a graph of a histogram of a fundus image.

FIG. 6 is a graph of a conversion table for dynamic range compression.

FIG. 7 is a graph of a histogram of a fundus image.

FIG. 8 is a graph of a conversion table for dynamic range compression.

FIG. 9 is a configuration diagram of a second embodiment.

FIG. 10 is an explanatory diagram of a fundus image.

FIG. 11 is an explanatory diagram of a fundus image.

FIG. 12 is a graph of a histogram of a fundus image.

FIG. 13 is a graph of a conversion table for dynamic range compression.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Observation light source 4 Imaging light source 7 Perforated mirror 10 Half mirror 11 Imaging means 12 Fixation target 14 Image memory 15 Image processing unit 17 TV monitor 18 Shooting operation control unit 19 Region of interest setting unit 21 Left and right eye detection unit 22 Fixation Position detector

Claims (6)

[Claims] An image sensor having a plurality of pixels for converting an eye image to be examined into an electric signal, a digital converter for converting the plurality of pixel outputs into digital signals, and a predetermined area pixel of the image sensor And first and second detecting means for respectively detecting a high level signal and a low level signal among the level signals indicating the luminance level of the digital signal in the above, and the digital signal based on the high level signal and the low level signal. Setting means for setting a signal compression ratio of the digital signal of the conversion unit. 2. The method according to claim 1, wherein the setting unit includes a compression ratio of a level signal by the digital conversion unit between the high level signal and the low level signal, and a compression ratio between the high level signal and the low level signal. 2. The ophthalmologic photographing apparatus according to claim 1, wherein the compression ratio of the level signal by the digital conversion unit is different. 3. The method according to claim 2, wherein the setting unit sets a compression ratio of a level signal by the digital conversion unit between the high level signal and the low level signal, the compression ratio being not between the high level signal and the low level signal. 2. The ophthalmologic photographing apparatus according to claim 1, wherein the compression ratio of the level signal by the digital conversion unit is reduced. 4. The predetermined pixel area for detecting the high level signal includes a nipple of the eye to be inspected, and the predetermined pixel area for detecting the low level signal includes a macula of the eye to be inspected. The ophthalmologic photographing apparatus according to claim 1, wherein: 5. The ophthalmologic photographing apparatus according to claim 1, further comprising a fixation target for guiding a fixation position of the subject's eye, wherein the predetermined pixel area is changed according to the fixation target. 6. The ophthalmologic photographing apparatus according to claim 1, wherein the predetermined pixel area is a predetermined fixed area.