KR102504865B1 - Binocular fundus camera - Google Patents

Abstract

The present invention provides a binocular imaging device. The binocular imaging device has a first optical axis twisted at a first angle with respect to the optical axis of the left eye, a left eye capture unit for imaging the pupil and retina of the left eye, and a second optical axis twisted at a second angle with respect to the optical axis of the right eye; The right eye capture unit for capturing the pupil and the retina of the right eye, the left eye capture unit and the right eye capture unit are moved in a first direction to align the optical axis of the left eye capture unit with the pupil of the left eye and set the optical axis of the right eye capture unit to the right eye A first direction movement mechanism for aligning the pupil of the eye, and a second direction movement mechanism for moving the left eye capture unit and the right eye capture unit in a second direction toward both eyes of the subject.
This patent application was carried out by the Ministry of Trade, Industry and Energy's Ophthalmic Optical Medical Device Globalization Support Project (IP strategy analysis and patent support, task number: B0081012000397, research period 2020. 12. 01 ~ 2021. 06. 30).

Description

Binocular imaging device {Binocular fundus camera}

The present invention relates to a binocular imaging device. This patent application was carried out by the Ministry of Trade, Industry and Energy's Ophthalmic Optical Medical Device Globalization Support Project (IP strategy analysis and patent support, task number: B0081012000397, research period 2020. 12. 01 ~ 2021. 06. 30).

The retina is the inner wall that surrounds the innermost part of the eye. In the retina, the macula is located in a portion corresponding to the visual axis and is the region with the best resolving power. In the retina, the blind spot (optical disc) is located in the direction of the nose from the macula, and is an area where optic nerve bundles connected to nerve cells located in the retina are gathered. On the other hand, the visual axis is tilted at a slight angle from the optical axis, which is a path of light entering vertically through the pupil. Because of this, the conventional retinal imaging device can take a retinal image after inducing the patient's line of sight to be distorted at a slight angle from the front.

It is intended to provide a binocular imaging device capable of simultaneously imaging the retinas of both eyes.

Meanwhile, it is intended to provide a binocular imaging device capable of simultaneously imaging the retinas of both eyes while guiding a patient's eyes to a specific direction and/or distance.

The present invention provides a binocular imaging device. The binocular imaging device has a first optical axis twisted at a first angle with respect to the optical axis of the left eye, a left eye capture unit for imaging the pupil and retina of the left eye, and a second optical axis twisted at a second angle with respect to the optical axis of the right eye; The right eye capture unit for capturing the pupil and the retina of the right eye, the left eye capture unit and the right eye capture unit are moved in a first direction to align the optical axis of the left eye capture unit with the pupil of the left eye and set the optical axis of the right eye capture unit to the right eye A first direction movement mechanism for aligning the pupil of the eye, and a second direction movement mechanism for moving the left eye capture unit and the right eye capture unit in a second direction toward both eyes of the subject. Here, the left eye capture unit and the right eye capture unit may be disposed to converge toward both eyes of the examinee.

In one embodiment, the left eye capture unit and the right eye capture unit simultaneously capture the left eye and the right eye, respectively, to generate a retinal image.

In one embodiment, the left eye photographing unit optically couples a ring-shaped white polarized light incident to the left eye along the first optical axis and the retinal illumination system, and photographs the pupil of the left eye to obtain a first tracking image. A pupil tracking optical system that generates a retinal tracking optical system that is optically coupled to the retinal illumination system, photographs the retina to generate a second tracking image, and detects white polarized light reflected from the retina of the left eye to generate a retinal image. A photographing optical system may be included.

In one embodiment, the second tracking image is used to track the position of the optical disc in the retina, and the second tracking image may be blurred.

In an embodiment, a focal position for photographing the optical disk on the retina may be determined by a focus value that is the sum of pixel value differences between pixels belonging to an optical disk correspondence area tracked by the left eye photographing unit.

In one embodiment, the focus value may be calculated after blurring the area corresponding to the optical disk.

In one embodiment, the focal position may be determined by a maximum value among focus values calculated in a plurality of optical disk corresponding regions acquired by photographing at a constant focal interval in one focal interval.

In an embodiment, the focal position may be determined by a maximum value among focus values calculated in a plurality of optical disk corresponding regions obtained by taking pictures at different focal intervals in a plurality of focal sections in which some sections overlap.

According to the present invention, the retinas of both eyes can be simultaneously photographed in a state in which the patient's eyes are directed forward. In particular, the retinas of both eyes may be simultaneously photographed while inducing the patient's eyes to face a specific direction or to look at a near or far object.

Hereinafter, the present invention will be described with reference to embodiments shown in the accompanying drawings. For ease of understanding, like reference numerals have been assigned to like elements throughout the accompanying drawings. The configurations shown in the accompanying drawings are only exemplary implemented embodiments to explain the present invention, and are not intended to limit the scope of the present invention thereto. Particularly, in the accompanying drawings, some of the elements represented in the drawings are somewhat exaggerated in order to aid understanding of the invention.
1 is a diagram schematically illustrating a binocular imaging apparatus according to an exemplary embodiment.
FIG. 2 is a diagram illustratively illustrating a state in which a photographing unit of the binocular imaging apparatus illustrated in FIG. 1 is viewed from above.
FIG. 3 is a view exemplarily illustrating a cross-section of a photographing unit of the binocular imaging apparatus illustrated in FIG. 2 .
4 is a diagram schematically illustrating a binocular imaging device according to another exemplary embodiment.
5 is a diagram illustrating gaze guidance using a virtual image as an example.
FIG. 6 is a diagram illustratively illustrating a state in which a photographing unit of the binocular imaging device illustrated in FIG. 4 is viewed from above.
FIG. 7 is a view exemplarily illustrating a cross-section of a photographing unit of the binocular imaging apparatus illustrated in FIG. 6 .
8 is a flowchart illustrating the operation of the binocular imaging device by way of example.
9 is a flowchart illustrating the pupil tracking illustrated in FIG. 8 by way of example.
10 is a diagram exemplarily illustrating a process of tracking a pupil in a first tracking image.
11 is a diagram for explaining an operation used in a tracking process by way of example.
12 is a diagram exemplarily illustrating a process of tracking a near-infrared reflection area in a first tracking image.
13 is a flowchart illustrating the retinal tracking illustrated in FIG. 8 as an example.
14 is a diagram exemplarily illustrating a process of tracking an optical disc in a second tracking image.
15 is a flowchart illustrating the auto focusing illustrated in FIG. 8 as an example.
16 is a flowchart illustrating an exemplary embodiment in which retina tracking and auto focusing are simultaneously performed.
17 is a diagram for illustratively explaining blur processing for accurately calculating a focus value.
18 is a flowchart illustrating auto focusing capable of reducing a time for calculating a focus value by way of example.
19 is a diagram for explaining auto focusing capable of reducing a time for calculating a focus value by way of example.
20 is a functional view of a binocular imaging device.

Since the present invention can have various changes and various embodiments, specific embodiments are illustrated in the drawings and will be described in detail through detailed description. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention. In particular, the functions, features, and embodiments described below with reference to the accompanying drawings may be implemented alone or in combination with other embodiments. Therefore, it should be noted that the scope of the present invention is not limited only to the forms shown in the accompanying drawings.

Meanwhile, among terms used in this specification, expressions such as “substantially”, “almost”, and “about” are expressions for considering margins applied in actual implementation or possible errors. Unless otherwise specified, "side" or "horizontal" refers to the left-right direction of the drawing, and "vertical" refers to the top-down direction of the drawing.

Throughout the appended drawings, the same or similar elements are referred to using like reference numerals.

1 is a diagram schematically illustrating a binocular imaging apparatus according to an exemplary embodiment.

The binocular imaging device 100 simultaneously captures the retinas of both eyes. To this end, the binocular imaging device 100 includes a left eye imaging unit 110L for imaging the retina of the left eye of both eyes of an examinee and a right eye imaging unit 110R for imaging the retina of the right eye. In the left eye capture unit 110L, the optical axis OA _cam of the left eye capture unit 110L is distorted at a first angle θ ₁ with respect to the direction perpendicular to the left eye, for example, the optical axis OA _eye of the left eye. . Similarly, the optical axis of the right eye photographing unit 110R is distorted at a second angle θ ₂ with respect to the optical axis of the right eye, eg, the optical axis of the right eye photographing unit 110R. Here, the absolute values of the first angle θ ₁ and the second angle θ ₂ may be substantially the same. That is, the left eye capture unit 110L and the right eye capture unit 110R may be arranged in a converged manner as they are closer to the subject. The pupil of the left eye is positioned on the optical axis of the left eye capture unit 110L, and the pupil of the right eye is positioned on the optical axis of the right eye capture unit 110R. Accordingly, the left eye capture unit 110L photographs the retina of the left eye through the pupil of the left eye, and the right eye capture unit 110R photographs the retina of the right eye through the pupil of the right eye.

The left eye capture unit 110L and the right eye capture unit 110R may each move in a first direction, eg, an x-axis direction. The distance d _eye between both eyes of the subject, eg, the distance between the pupils, may be different for each subject. In the case of the existing equipment that separately scans both eyes, the distance d _eye between both eyes did not need to be considered. However, in order to capture both eyes simultaneously, the optical axes of the left eye capture unit 110L and the right eye capture unit 110R must be aligned with the pupils of both eyes. Accordingly, the binocular imaging apparatus 100 includes a left eye movement mechanism 130L for moving the left eye capture unit 110L in the first direction and a right eye movement mechanism 130R for moving the right eye capture unit 110R in the first direction. may contain at least one. The left eye movement mechanism 130L and the right eye movement mechanism 130R are collectively referred to as the first direction movement mechanism 130 . The movement mechanism in the first direction may be implemented with known devices such as motors and actuators.

Meanwhile, the left eye capture unit 110L and the right eye capture unit 110R may move together in the second direction, eg, the y-axis direction. The binocular imaging device 100 includes a housing 150 accommodating a left eye capture unit 110L, a right eye capture unit 110R, and a moving mechanism in the first direction. In one embodiment, the left eye capture unit 110L and the right eye capture unit 110R may move forward or backward in the second direction inside the housing 150 . In another embodiment, when the housing 150 moves forward or backward in the second direction, the left eye capture unit 110L and the right eye capture unit 110R may move forward toward or backward from the eyes of the subject. Hereinafter, it is assumed that the second direction movement mechanism moves the left eye capture unit 110L and the right eye capture unit 110R in the second direction, encompassing the two embodiments. The second direction movement mechanism may be implemented with known devices such as motors and actuators.

The second direction movement mechanism 140 may move at different speeds depending on the distance from the subject. The movement distance per unit time of the second direction movement mechanism 140, that is, the movement speed, is the highest between the initial position H where the left eye capture unit 110L and the right eye capture unit 110R are farthest from the subject and a certain position p ₁ . . Between p ₁ and a position p ₂ suitable for retinal imaging, the moving speed decreases. Finally, when the left eye capture unit 110L and the right eye capture unit 110R reach p ₂ , the second direction movement mechanism 140 stops the movement of the left eye capture unit 110L and the right eye capture unit 110R. . Here, the left eye capture unit 110L and the right eye capture unit 110R may be moved in the first direction by the first direction movement mechanism 130 . In other words, while the left eye capture unit 110L and the right eye capture unit 110R are being moved by the second direction movement mechanism 140 , the position of the pupil may be detected and the optical axis may be aligned. Additionally or alternatively, the left eye capture unit 110L and the right eye capture unit 110R may move forward or backward while capturing the retinas of both eyes. By varying the movement speed, a movement time for the binocular imaging apparatus 100 to a position suitable for retinal imaging may be reduced. In particular, the human eye tends to be easily fatigued by light, so the effect of fatigue on the eye can be minimized by imaging the retina while moving.

FIG. 2 is a view illustratively illustrating a state in which a photographing unit of the binocular imaging apparatus illustrated in FIG. 1 is viewed from above, and FIG. 3 is a diagram exemplarily illustrating a cross-section of a photographing unit of the binocular imaging apparatus illustrated in FIG. 2 . Since the left eye capture unit 110L and the right eye capture unit 110R have the same structure except for the state in which they are arranged toward both eyes, hereinafter, they will be collectively referred to as the capture units 110L and 110R without distinction.

Maculae 12L and 12R and blind spots 13L and 13R are located on the retinas of both eyes 10L and 10R. Looking at the photographs 10La and 10Ra of the retinas of both eyes, it can be seen that the structures of both eyes 10L and 10R are substantially symmetrical. The distance between the blind spots 13L and 13R is shorter than the distance between the maculas 12L and 12R. An optical axis OA _eye , which is a path of light vertically passing through the pupils 11L and 11R, extends between the maculas 12L and 12R and the blind spots 13L and 13R. On the other hand, the visual axis, which is a path for light passing through the pupils 11L and 11R to reach the maculas 12L and 12R, is twisted at a predetermined angle with respect to the optical axis OA _eye , so that the visual axis of both eyes 10L and 10R crosses in front of the subject.

Referring to FIGS. 2 and 3 together, the photographing units 110L and 110R include a retinal illumination system in which a pupil tracking optical system and a retina imaging optical system are optically coupled. The retinal illumination system may be implemented in various structures. The optical axis of the retinal illumination system is the optical axis OA _cam of the photographing units 110L and 110R, and the pupil tracking optical system composed of the first illumination 200 and the first camera 210 is optically coupled to the retinal illumination system by a beam splitter. . Meanwhile, the retinal imaging optical system includes a mirror 230, a polarizer 231, and a second camera 232, and is also optically coupled to the retinal illumination system by the beam splitter 203. The illustrated structure is only an example, and the location of the camera and lighting may be arranged differently, of course.

The retinal illumination system inputs ring-shaped white polarized light for retinal imaging into the eyes 10L and 10R. The retinal illumination system is optically coupled with the pupil tracking optical system and the retinal imaging optical system. Accordingly, the pupil tracking optical system and the retinal imaging optical system may externally radiate light along at least a part of the optical axis OA _cam of the photographing units 110L and 110R, or may receive light reflected from the cornea or retina. To this end, the retinal illumination system includes a white light illumination 220, a filter 222, an annular aperture 223, a polarizer 224, a plurality of beam splitters 201, 202, 203 arranged along the optical axis OA _cam , and a plurality of collimating lenses 204 and 221.

In the illustrated structure, a white light illumination 220 generating white light is disposed at one end farthest from the eye along the optical axis OA _cam . The other end of the retinal illumination system may be opened so that light may be incident to the eye or light reflected from the eye may be incident to the inside. Straightness of the white light generated by the white light illumination 220 may be improved by the first collimating lens 221 . A filter 222 , an annular aperture 223 and a polarizer 224 are disposed in front of the first collimating lens 221 . White light has a ring shape by the annular aperture 223 and passes through the polarizer 224 to become ring-shaped white light polarization. The ring-shaped white polarized light minimizes reflection by the cornea, thereby improving the quality of a photographed retinal picture. The first beam splitter 201 passes ring-shaped white polarized light traveling toward the eye along the optical axis OA _cam , and refracts light generated by the first illumination 200, for example, near infrared rays toward the eye. . The second beam splitter 202 is disposed in front of the first beam splitter 201, passes ring-shaped white polarized light and near infrared rays traveling toward the eye, and transmits the near infrared rays reflected from the eye to the first camera 210. bend towards The third beam splitter 203 is disposed in front of the second beam splitter 202, passes ring-shaped white polarized light and near infrared rays traveling toward the eye, passes near infrared rays reflected from the eye, and reflects from the eye. The white polarized light is refracted toward the second camera 232 .

The pupil tracking optical system includes a first light 200 optically coupled to the retinal illumination system by a first beam splitter 201 and a first camera 210 optically coupled to the retinal illumination system by a second beam splitter 202. ) may be included. The first illumination 200 generates near-infrared rays, and the first camera 210 may detect the near-infrared rays reflected from the eye 10L or 10R. Although not shown, the first camera 210 may include auto focus.

The retinal imaging optical system may include a mirror 230 , a polarizer 231 , and a second camera 232 . The third beam splitter 203 refracts white polarized light reflected from the retina toward the mirror 230 . The mirror 230 refracts the white polarized light refracted by the third beam splitter 203 toward the second camera 232 . The white polarized light refracted by the mirror 230 passes through the polarizer 231 and reaches the second camera 232 . Although not shown, the second camera 232 may include auto focus.

Now, pupil tracking and retinal imaging will be schematically described with reference to FIG. 3 .

In a1 , while or after the photographing units 110L and 110R reach a distance suitable for retinal imaging, the first illumination 200 is turned on. The first illumination 200 emits near-infrared rays required for pupil tracking. Near-infrared rays are useful for acquiring a first tracking image having high contrast despite reflection by the cornea. Also, unlike visible light, since it causes less eye fatigue, it can be continuously irradiated toward the eye during the pupil tracking process.

In a2 , after the first light 200 is turned on, the first camera 210 detects near-infrared rays reflected from the eye to generate a first tracking image. The first tracking image is used to track the pupil. When the pupil position is determined, the first directional movement mechanism moves the photographing units 110L and 110R to a position where the optical axis OA _cam of the photographing parts 110L and 110R passes through the pupil at an angle and extends to the retina (hereinafter, the optimal pupil). position) in the first direction.

At b1, if the pupil is at the optimal pupil position as a result of pupil tracking, the white light illumination 220 is turned on. The white light illumination 220 becomes ring-shaped polarized white while passing through the retinal illumination system, and enters the eye through a partial region of the pupil.

In b2, after the white light illumination 220 is turned on, the second camera 232 detects the white polarized light reflected from the retina and simultaneously generates retinal images of both eyes 10L and 10R. Here, the first direction movement mechanism may move the photographing units 110L and 110R generating retinal images in the first direction, respectively.

4 is a diagram schematically illustrating a binocular imaging device according to another embodiment, and FIG. 5 is a diagram exemplarily illustrating gaze guidance using a virtual image. Here, a description of the same configuration as that shown in FIG. 1 will be omitted, and differences will be mainly described.

Referring to FIG. 4 , the both eye imaging device 101 simultaneously captures the retinas of both eyes 10L and 10R while guiding the gaze of the subject. The binocular imaging apparatus 101 includes a left eye imaging unit 111L for imaging the retina of the left eye of both eyes of an examinee, a right eye imaging unit 111R for imaging the retina of the right eye, a left eye image display system for displaying a left eye image, and a right eye image. It includes a right eye image display system that displays. The left eye capture unit 111L is distorted at a first angle θ ₁ with respect to a direction perpendicular to the left eye 10L, and the right eye capture unit 111R is disposed at a second angle with respect to a direction perpendicular to the right eye 10R. The optical axis is arranged to be twisted at θ ₂ . In contrast, the left eye image display system is arranged so that the left eye image is displayed in front of the left eye, and similarly, the right eye image display system may be arranged so that the right eye image is displayed in front of the right eye. In summary, the optical axis OA _VR of the image display system is different from the optical axis OA _cam of the photographing units 111L and 111R for photographing the retina at a predetermined angle.

The images displayed by the left eye image display system and the right eye image display system are recognized as a single image by the subject. Therefore, hereinafter, these two are collectively referred to as virtual images without distinction. The virtual image is used to direct the subject's eyes to a specific location. That is, through gaze induction using a virtual image, the subject's pupil may be moved to an optimal pupil position or the pupil may be moved to a position suitable for retinal imaging (hereinafter, referred to as an optimal imaging position). The virtual image may be a 2D or 3D image. The screens 300, 300a, 300b, and 300c are examples of virtual images viewed by the subject. One or more objects 301 and 302 may be displayed on the screens 300, 300a, 300b, and 300c. The first object 301 represents the optimal pupil position and/or the optimal photographing position, and the second object 302 represents the subject's actual pupil position. For example, the display position of the first object 301 on the screen 300 is determined by the direction in which the photographing units 111L and 111R are directed (or the direction in which the optical axis OA _cam is directed), and the second object 302 is A display position on the screen 300 may be determined by the subject's actual pupil position.

As an example, the first object 301a may be used to move the subject's gaze in a third direction, for example, in the z-axis direction. The photographing units 111L and 11R may track the subject's pupil while moving in the x-axis and/or y-axis directions, or may track a specific object on the retina, for example, an optical disc, in a stationary state. For example, the photographing units 111L and 11R may move in the z-axis direction by a third movement mechanism (not shown), but in order to finely move the subject's pupil in the z-axis direction, the movement mechanism 130, 140) may be more effective than pupil movement through gaze induction. In view 300a, the optimal pupil position 301a is above the actual pupil position 302a. When the subject moves the pupil upward while viewing the virtual image 300a displayed by the left eye image display system and the right eye image display system, the binocular imaging device 101 moves the second object 302a upward through pupil tracking. indicate as When the first object 301a and the second object 302a overlap, the binocular imaging apparatus 101 may perform retinal imaging by determining that the subject's pupil position has reached an optimal pupil position or an optimal imaging position.

As another embodiment, the first object 301b may be used to guide the subject's gaze in a first direction, for example, in an x-axis direction. In a state in which the photographing units 111L and 111R are stopped, when the subject moves the pupil, the photographing units 111L and 111R may photograph various regions of the retina. In the screen 300b, the optimal shooting position 301b is to the right of the actual pupil position 302b. When the examinee moves the pupil to the right while viewing the virtual image 300b displayed by the left eye image display system and the right eye image display system, the binocular imaging device 101 moves the second object 302b to the right through pupil tracking. indicate as When the first object 301b and the second object 302b overlap, the binocular imaging apparatus 101 may perform retinal imaging by determining that the subject's pupil position has reached the optimal imaging position.

Referring to FIG. 5 together, in another embodiment, the virtual images displayed by the left eye image display system and the right eye image display system may be 3D images having perspective. The 3D image may induce the subject's gaze or dilate the pupil so that the angle between the gaze narrows or widens. In the human eye, the angle between the gaze narrows and the pupil dilates to receive a relatively large amount of light in order to see a distant object. Conversely, to see close objects, the angle between the gazes widens and the pupil narrows, allowing in relatively little light.

In the 3D image having a sense of perspective, the pupil may be expanded or contracted according to the perceived distance of the subject. When the pupil is dilated, a large area of the retina can be imaged. In the screen 300c, the optimal pupil position or optimal imaging position 301c is farther from the actual pupil position 302c. When the examinee moves the viewpoint to the first object 301c while viewing the virtual image 300c displayed by the left eye image display system and the right eye image display system, the binocular imaging device 101 recognizes pupil dilation through pupil tracking and detects pupil dilation. 2 It is indicated that the object 302c moves toward the first object 301c.

Meanwhile, in a 3D image having a sense of perspective, the angle between gazes may be increased or decreased according to a distance perceived by the subject. The second objects 302L' and 302R' displayed by the left eye image display system and the right eye image display system allow the subject to perceive that the second object 302c' is located relatively close. As a result, the angle between the eyes of the examinee is widened. Meanwhile, the second objects 302L″ and 302R″ displayed by the left eye image display system and the right eye image display system allow the subject to recognize that the second object 302c″ is located relatively far away. As a result, the angle between the eyes of the examinee is narrowed.

The above-described embodiments may be individually implemented or combined. For example, when the pupil is moved in the first direction or the third direction, the first object may be displayed in a 3D image through which the examinee can feel perspective. As another example, the pupil may move to an optimal pupil position by a combination of movement in the first to third directions.

Meanwhile, as the area of the first object becomes smaller, the subject's pupil may be more accurately guided to the optimal pupil position. Even if the pupil is sufficiently dilated, due to the anatomy of the eye, a particular location of the pupil may be optimal for retinal imaging. Accordingly, the binocular imaging device 101 may guide the gaze of the subject so that the specific position of the pupil is located on the optical axis OA _cam of the imaging units 111L and 111R.

FIG. 6 is a view illustratively illustrating a state in which a photographing unit of the binocular imaging apparatus illustrated in FIG. 5 is viewed from above, and FIG. 7 is a diagram exemplarily illustrating a cross-section of a photographing unit of the binocular imaging apparatus illustrated in FIG. 6 . Descriptions of the same components as those shown in FIGS. 2 and 3 will be omitted, and differences will be mainly described.

Referring to FIGS. 6 and 7 together, the photographing units 111L and 111R include a retinal illumination system in which a pupil tracking optical system and a retina tracking/photography optical system are optically coupled. The retina tracking/photography optical system includes a second illuminator 400, a mirror 230, a polarizer 231, and a second camera 232, and is optically coupled to the retinal illuminator by a beam splitter 203. . The retina tracking/photography optical system tracks the optical disk and photographs the retina to generate a retinal image. The illustrated structure is only an example, and the location of the camera and lighting may be arranged differently, of course.

The imaging units 111L and 111R include a left eye image display system and a right eye image display system (hereinafter collectively referred to as image display systems) disposed in the retinal illumination system. The image display system may include a display 410, a fourth collimating lens 420, and a fourth beam splitter 430. The fourth beam splitter 430 of the image display system is disposed on the optical axis OA _cam of the retinal illumination system, but the optical axis OA _VR of light refracted by the fourth beam splitter 430 is not parallel to the optical axis OA _cam of the retinal illumination system. . That is, the optical axis OA _VR is distorted by a third angle θ ₃ with respect to the optical axis OA _cam . Here, the absolute value of the third angle θ ₃ may be substantially the same as the absolute value of the first angle θ ₁ or the second angle θ ₂ . Due to this, the virtual image output by the video display system is displayed in front of the left eye and the right eye, respectively.

The image display system may be disposed at the opposite end of the white light illumination 220, that is, at the other end closest to the eye. The second illumination 400 may radiate near infrared rays toward the eyes. The display 410 outputs a virtual image to be displayed on both eyes of the subject. The image is collected by the fourth collimating lens 420 and then refracted by the fourth beam splitter 430 to enter the eye. Meanwhile, light traveling along the optical axis OA _cam of the retinal illumination system passes through the fourth beam splitter 430 .

Now, referring to FIG. 7, retinal tracking will be schematically described.

In c1, after the photographing units 111L and 111R reach a distance suitable for retinal imaging, a virtual image for guiding the subject's gaze is output. By the pupil movement, the optical disk on the retina may correspond to the photographing units 111L and 111R.

At c2, the second light 400 is turned on. The second illumination 400 emits near infrared rays necessary for tracking the position of the optical disk. Near-infrared rays are useful for obtaining a second tracking image having high contrast despite reflection by the cornea. In addition, unlike visible light, since it causes less eye fatigue, it can be continuously irradiated toward the eye during the retinal tracking process.

In c3 , after the second illumination 400 is turned on, the second camera 232 detects the near-infrared rays reflected from the retina and generates a second tracking image. The second tracking image is used to track the optical disc. If the optical disc is not in the proper position, the virtual image is changed.

8 is a flowchart illustrating the operation of the binocular imaging device by way of example.

In one embodiment, the binocular imaging apparatus 101 described with reference to FIGS. 4 to 7 may sequentially perform pupil tracking (S10), retina tracking (S20), auto focusing (S30), and retinal imaging (S40). can The pupil tracking (S10) is a process of tracking the subject's pupil. By pupil tracking, the pupils may come to an optimal pupil position by moving the photographing units 111L and 111R in the first direction or by changing the gaze or posture of the subject. Retina tracking (S20) is a process of tracking the optical disc on the retina. That is, the subject's gaze is guided based on the measured position of the optical disc on the retina. Through the retinal tracking (S20), the subject may move the pupil to an optimal photographing position. Additionally, in the network tracking process ( S20 ), the position of the area where the second light 400 is reflected (hereinafter referred to as a reflection area) may be tracked. The reflection area may be used to adjust the distance between the photographing units 111L and 111R and the pupil. Auto focusing (S30) is a process of adjusting the focus to photograph the optical disk. The retinal imaging (S40) is a process of generating a retinal image by imaging the optical disk. Meanwhile, the binocular imaging device 100 described with reference to FIGS. 1 to 3 may sequentially perform part of pupil tracking ( S10 ), auto focusing ( S30 ), and retinal imaging ( S40 ).

As another embodiment, the binocular imaging apparatus 101 described with reference to FIGS. 4 to 7 may simultaneously perform retina tracking ( S20 ) and auto focusing ( S30 ) after performing pupil tracking ( S10 ). Retinal tracking ( S20 ) may be performed while changing the focus of the second camera 232 using auto focus.

Each process is described in detail below.

FIG. 9 is a flowchart exemplarily illustrating the pupil tracking operation illustrated in FIG. 8, FIG. 10 is a diagram exemplarily illustrating a process of tracking a pupil in a first tracking image, and FIG. It is a diagram for explaining the calculation by way of example, and FIG. 12 is a diagram showing a process of tracking a near-infrared reflection area in the first tracking image as an example.

9 to 12 together, the photographing units 111L and 111R are moved to their initial positions by the moving mechanisms 130 and 140 (S100). The initial position is a position suitable for retinal imaging and can be determined using a model eye. Meanwhile, the initial position may also be determined from positions of the photographing units 111L and 111R in cases where the retina is clearly photographed. The moving mechanisms 130 and 140 move the capturing units 111L and 111R in the first direction and/or the second direction so that the capturing units 111L and 111R are aligned in their initial positions.

An image for guiding the gaze of the subject in a third direction, for example, up and down, is output (S110). The pupil is moved by gaze guidance. An image used for pupil tracking may be a 2D image. The output image is refracted by the fourth beam splitter 430, which is twisted at a third angle θ ₃ with respect to the retinal illumination system and directed toward the front of both eyes, and proceeds toward the eyes. Meanwhile, step S110 may be selectively performed.

The first camera 210 captures the eye and generates a first tracking image 500 (S120). The first light 200 may be turned on during the pupil tracking process to radiate near-infrared rays toward the eyes. The first camera 210 generates a first tracking image 500 by detecting near-infrared rays reflected from around the eye including the pupil. The first tracking image 500 includes a reflection area 502 indicating a location where near infrared rays are reflected by the cornea and/or iris, and an iris 503 where the near infrared rays are reflected.

A pupil is detected in the first tracking image 500 (S130). Referring to FIG. 10 , since the pupil passes near-infrared rays, the reflected near-infrared rays are relatively small, whereas the iris 503 around the pupil reflects the near-infrared rays relatively well. An effective area 504 for determining an optimal pupil position may be defined in the first tracking image 500 . The first camera 200 may generate a black and white image or a color image. In the case of a black-and-white image, each pixel of the first tracking image 500 has a gray scale value, and in the case of a color image, each pixel of the first tracking image 500 has red, green, and blue values. In the case of a black-and-white image, a histogram 505 of the effective area 504 is generated using gray scale as a pixel value, and in case of a color image, a blue value as a pixel value. The horizontal axis of the histogram 505 represents the pixel value, and the vertical axis represents the number of pixels.

Since very little near-infrared light is reflected, the pupil is a relatively dark region. Therefore, in the histogram 505, relatively small pixel values displayed on the left represent the pupil, and relatively large pixel values displayed on the right represent the remaining area surrounding the pupil, such as the iris. The boundary between the pupil and the iris may be expressed as a pixel value in which a relatively small pixel value rapidly decreases. Accordingly, a first valley 506 in which a small pixel value rapidly decreases can be detected, and a pupil can be detected using a pixel value corresponding to the valley 506 .

For example, the first valley 506 can be determined as follows. First, the number G _y of pixels having any one pixel value P _x is less than or equal to the number of pixels G _y1 having a small pixel value of 1 to n (n is 2 or more), and a pixel value 1 to n larger If the number of pixels having is less than or equal to G _y2 , the corresponding pixel value P _x is selected. This condition is a condition for detecting most valleys existing on the histogram 505. If more than one pixel value P _x is detected, the number G _y of pixels having the detected pixel value P _x is K times the maximum of the number of pixels having a pixel value smaller than the detected pixel value P _x (where K is less than 1), a pixel value P _x smaller than 1 is selected. This condition is a condition for finding a valley past a peak. If two or more pixel values P _x are detected, the pixel value P _x whose maximum value is greater than M ₁ among the number of pixels having a pixel value smaller than the detected pixel value P _x is selected. Here, M ₁ may be determined according to the total number of pixels included in the effective area 504 . For example, if the total number of pixels included in the effective area 504 is 80,000, M ₁ may be 120 corresponding to about 0.15%. This condition is a condition for finding a valley past a relatively large peak.

The pupil image 507 before noise removal shows a region 508 having a pixel value equal to or smaller than the determined pixel value of the first valley 506 . In addition to the reflective region 502 formed by the cornea and/or iris located around the pupil, a region having a large pixel value, ie, noise, may exist inside the pupil. Noise may be removed from the pupil image 507 before noise removal by applying a morphology operation. 11 illustrates an image processing effect by morphology calculation.

Erode shrinks bright areas, that is, areas with relatively large pixel values. Therefore, by the erosion process, regions with a small area corresponding to the noise are removed.

Dilate expands the bright area. Accordingly, a dark area located within a bright area, that is, an area having a relatively small pixel value is removed.

Open applies dilation after applying erosion. According to the opening, noise having a small area is removed, and areas reduced by erosion are expanded again. That is, opening is an operation for removing noise.

Close applies dilation followed by erosion. By closing, the dark regions located inside the bright regions are removed, and the bright regions expanded by dilation are contracted. That is, closing is an operation for removing dark areas inside bright areas.

The morphology closing operation and opening operation are applied to the pupil image 507 before denoising. The size of the filter applied during the closing operation and the opening operation may be larger than filters applied in other operations below because the area 508 to be filtered is relatively wide.

The pupil image 508 after noise removal includes a pupil area 509 . In the image illustrated in FIG. 10, the pupil area 509 is normally located within the effective area 504, but in the pupil tracking process, an area similar to the pupil area 509 may be located in the effective area 504, Only a part of the area 509 may be located in the effective area 504 . Among the filtered regions, a region corresponding to the pupil is determined using an anatomical characteristic of the pupil, for example, a width/height ratio of the pupil. Then, in the first tracking image, the location and size of the pupil area 509 are calculated.

Meanwhile, the first tracking image 500 may be used to measure the distance between the subject's eyes and the photographing units 111L and 111R. Referring to FIG. 12 , there may be a plurality of first lights 200 , or near-infrared rays generated by a single first light 200 may reach the eye through a plurality of separated light paths. Near infrared rays reflected from the vicinity of the cornea and/or iris may be detected by the first tracking image 500 . From the position of the reflective region 502 and/or the distance between regions, not only the position of the pupil but also the distance between the subject's eye and the photographing units 111L and 111R can be measured. The distance between the subject's eyes and the photographing units 111L and 111R may be adjusted by the measured distance.

To this end, a reflection tracking image 510 including at least a part of the effective region 504 is extracted from the first tracking image 500 . A histogram 511 of the reflection tracking image 510 is generated. The horizontal axis of the histogram 511 represents the pixel value, and the vertical axis represents the number of pixels.

Contrary to the pupil, the reflective region 502 is a very bright region. Accordingly, in the histogram 511, the largest pixel values displayed on the right side represent the reflective region 502. The circumference of the reflective region 502 may be expressed as a pixel value in which the number of pixels rapidly decreases. For example, if the number G _y of pixels having any one pixel value P _x is smaller by M ₂ than the number of pixels G _y3 having pixel values greater than 1 to n (n is 2 or more), the pixel value P _x is determined around the reflective area 502. A large pixel value P _x is selected. Here, M ₂ may be determined according to the total number of pixels included in the reflection tracking image 510 .

The sub tracking image 513 before noise removal represents a candidate region 514 having the determined pixel value P _x . Noise may be removed from the sub-tracking image 513 before noise removal by sequentially applying morphology erosion and dilation operations. Since noise is relatively smaller than noise included in other images, the size of a filter used for erosion calculation may be relatively smaller than in other cases.

The sub tracking image 514 after noise removal may include a plurality of candidate regions 514 . Therefore, in order to determine the reflective region 515 associated with the near-infrared rays emitted from the photographing units 111L and 111R, conditions such as the area of each candidate region 513, the location of the center, and the distance between centers may be used.

It is determined whether the pupil area 509 is at the optimal pupil position (S140). For example, the determination may be made based on an overlapping ratio between the pupil area 509 and the region defined as the optimal pupil position or a distance between center points.

If the pupil area 509 is not at the optimal pupil position, the photographing units 111L and 111R may be moved in the first direction or a virtual image for guiding gaze up and down may be changed (S150). The changed virtual image reflects the pupil area 509 detected in the previous step. When the virtual image is changed, the examinee may adjust the gaze or posture while viewing the changed virtual image. When the pupil area 509 is at the optimal pupil position, the retinal tracking process (S20) begins.

FIG. 13 is a flowchart exemplarily illustrating the retinal tracking operation illustrated in FIG. 8 , and FIG. 14 is a diagram exemplarily illustrating a process of tracking an optical disk in a second tracking image.

A virtual image for guiding the gaze of the subject is output (S300). In the binocular imaging device 101 , the display 410 outputs the virtual image 300 illustrated in FIG. 4 . The output virtual image includes a first object 301 indicating an optimal capturing position and a second object 302 indicating an actual pupil position. The virtual image used for retinal tracking may be a 2D or 3D image. The output virtual image is twisted at a third angle θ ₃ with respect to the retinal illumination system and refracted by the fourth beam splitter 430 toward the front of both eyes, and then proceeds toward the eyes.

When the second light 400 is turned on, the second camera 232 captures the retina to generate a second tracking image 520 (S310). The second light 400 may be turned on during the retina tracking process to radiate near-infrared rays toward the eye. The second camera 232 generates a second tracking image 520 by detecting near-infrared rays reflected from the eye including the retina.

The overall brightness of the second tracking image 520 is compared with a predetermined threshold value (S320). While retina tracking (or autofocus) is in progress, if the subject closes his/her eyes, the second camera 232 cannot capture the retina, and since the second camera 232 detects only near-infrared rays reflected on the eyelids, The overall brightness of the tracking image 520 becomes very large. In this case, the step of detecting the optical disc in the second tracking image may be omitted. If the total brightness is greater than the threshold, the process returns to step S310, and if less than the threshold, the process proceeds to step S330.

The position of the optical disc 522 is detected from the second tracking image 520 (S330). Optical disc detection is similar to pupil detection or reflective area detection described above. In FIG. 20 , a histogram is generated for the valid region 521 of the second tracking image 520 . Optionally, in order to shorten the tracking time, a histogram of an effective region of an image obtained by reducing the second tracking image 520 at a predetermined ratio may be generated. The effective area 521 may be defined in the second tracking image 520 or a reduced second tracking image.

The optical disc 522 is a relatively bright area compared to other areas. Accordingly, in the histogram, pixels having relatively large pixel values may represent the optical disk 522 . Starting from the largest pixel value and going in a direction in which the pixel value decreases, a pixel value at which the number of pixels rapidly increases is determined. For example, if the number of pixels G _y having any one pixel value P _x is greater than the number of pixels G _y4 having a pixel value larger than 1 to n (n is 2 or more) by M ₃ , the pixel value P _x is It is determined by the circumference of the optical disk 521. Here, M ₃ may be determined according to the total number of pixels included in the effective area 521 .

In the second tracking image 520, regions having pixel values greater than the determined pixel values are determined. The optical disc image 522 before noise removal shows a region 523 having a pixel value greater than the determined pixel value. In the retina, not only the optical disk, but also a region having similar brightness, ie, noise, may exist. Noise may be removed from the optical disc image 522 before noise removal by applying a morphology operation. Two or more filtering operations, eg, morphology closing operation and morphology opening operation, are performed on the determined regions. Among the filtered regions, regions having a specific area are selected. Partial images are extracted from the center of the selected regions and blurred. A total sum of pixel value differences of pixels adjacent to the blurred image is calculated. The area with the largest calculated value is the optical disk 522 . The optical disc image 524 after noise removal includes an area 525 corresponding to the optical disc.

Based on the optical disc correspondence area 526, it is determined whether the pupil is at an optimal photographing position (S340). The optical disc corresponding area 526 may be a rectangular area including the optical disc. For example, the determination may use the position of the optical disc in the retinal image.

If the pupil is not at the optimal photographing position, a virtual image for guiding the gaze may be changed (S350). The changed virtual image reflects the position of the optical disk correspondence area 526 . When the virtual image is changed, the examinee may adjust the gaze while viewing the changed virtual image.

When the actual pupil position is at the optimal pupil position, auto focusing is performed (S360).

15 is a flowchart illustrating the auto focusing illustrated in FIG. 8 by way of example, and shows an embodiment in which retina tracking and auto focusing are sequentially performed.

The focus by auto focus is at the initial position (S400). Autofocus moves the focus formed by a plurality of lenses step by step from an initial position to a final position. The focal position changes according to the distance between the lenses, and the distance between the lenses is adjusted by a motor. Therefore, the focal position may be expressed as the rotation number and/or the rotation angle of the motor rotation shaft.

The second camera 232 photographs the optical disk to generate a retinal image (S410). At this time, the actual pupil position is at the optimal photographing position. The second camera 232 may photograph the retina while sequentially moving the focal point.

In the photographed image, the position of the optical disc is tracked in the retinal image (S420). Position tracking of the optical disk may be substantially the same as step S330 of FIG. 19 described above.

A focus value is calculated in the optical disc corresponding area 526 (S430). The focus value is the total sum of pixel value differences between pixels belonging to the optical disk correspondence area 526, and the focus value is related to the focus position when the image is captured. Compared to a relatively small focus value, a relatively large focus value indicates that the focus by autofocus is at a relatively good position. Accordingly, a clearer retinal image may be obtained.

It is determined whether the focus by auto focus has changed from the initial position to the final position (S440). Assuming that the number of rotations of the motor rotation shaft is 0 when the focus is at the initial position and the number of rotations of the final position is 100, if the measured number of rotations is greater than 0 and less than 100, the process proceeds to step S450, and the measured number of rotations is 100. If not, the process may proceed to step S460.

If the focus is between the initial position and the final position, the focus is changed by auto focus (S450).

When the focus is at the final position, auto focus is adjusted so that the focus is at the focus position having the largest focus value among the stored focus values (S460).

Thereafter, the white light illumination 200 is turned on to photograph the retina (S40).

16 is a flowchart illustrating an exemplary embodiment in which retina tracking and auto focusing are simultaneously performed.

A virtual image for guiding the gaze of the subject is output (S500), and the focus by auto focus is at an initial position (S510). Here, the order of step S500 and step S510 may be changed or performed simultaneously.

When the second light 400 is turned on, the second camera 232 captures the retina to generate a second tracking image 520 (S520). If the total brightness of the second tracking image 520 is greater than the threshold value (S530), the second camera 232 photographs the retina without changing the focus.

The position of the optical disc is detected in the second tracking image 520, and a focus value is calculated using the area 526 corresponding to the optical disc (S540). The position of the optical disk can be detected through the process described in step S330 of FIG. 13, and the focus value can be detected through the process described in step S430 of FIG. 15.

Based on the detected position of the optical disk, it is determined whether the pupil is at an optimal photographing position (S550). If it is not an optimal photographing position, a virtual image for guiding the gaze of the subject may be changed (S560). If it is the optimal photographing position, it is determined whether the focus by auto focus has changed from the initial position to the final position (S570). If the focus is between the initial position and the final position, the focus is changed by auto focus (S580). When the focus is at the final position, auto focus is adjusted so that the focus is at the focus position having the largest focus value among the stored focus values (S590). Thereafter, the white light illumination 200 is turned on to photograph the retina (S40).

17 is a diagram for illustratively explaining blur processing for accurately calculating a focus value.

The focus value is the total sum of pixel value differences between pixels belonging to the optical disk corresponding area 526, and a sharper retinal image can be created as the focus value increases. Therefore, the focus value of the correctly focused retinal image must be greater than the focus value of the incorrectly focused retinal image. However, there may be a case where a focus value of an incorrectly focused retinal image is greater than a focus value of an accurately focused retinal image. This is because the retinal image is a dark image generated by photographing a dark region inside the eye using near-infrared rays, and thus is more greatly affected by noise. Therefore, it is necessary to remove noise from the optical disc corresponding area 526.

A difference 541 between pixel values located on a straight line displayed in the retinal image 540 shows a state before noise is removed. An incorrectly focused retinal image contains more noise than a correctly focused retinal image. Since noise increases the difference between pixel values, the focus value of the retinal image increases as the amount of noise increases. To reduce noise, the entire retinal image 540 or the region 526 corresponding to the optical disc is blurred. In the retinal image 540, a specific region, eg, the optical disc and its surrounding blood vessels, is relatively more important, and other regions are relatively less important. In addition, the time required to blur the entire retinal image 540 having a relatively large area may be greater than the time required to blur the optical disc corresponding region 526 having a relatively small area.

Blur processing can use various blur filters. The blur filter may be, for example, any one of an average blur filter 542, a Gaussian blur filter 543, and a median blur filter. Here, the size of the blur filter is n x n (n is the number of pixels), and a pixel located at the center of the blur filter is referred to as a center pixel, and pixels located around the center pixel are referred to as peripheral pixels. The average blur filter 542 calculates an average value of pixel values of the center pixel and neighboring pixels, and uses the calculated average value as the pixel value of the center pixel. The Gaussian blur filter 543 applies different weights to the pixel values of the central pixel and the peripheral pixels, calculates an average value of the weighted pixel values, and takes the calculated average value as the pixel value of the central pixel. In the median blur filter, the median value of the pixel values of the central pixel and the peripheral pixels is set as the pixel value of the central pixel.

A difference 545 between pixel values located on a straight line displayed in the blurred retinal image 544 shows the state well after noise is removed. It can be seen that a large amount of noise is removed by the blur processing. Blur processing may be applied not only to calculate a focus value, but also to detect an optical disk during retinal tracking.

18 is a flowchart exemplarily illustrating auto focusing capable of reducing the time for calculating a focus value, and FIG. 19 is a diagram for exemplarily illustrating auto focusing capable of reducing a time for calculating a focus value.

Referring to FIGS. 18 and 19 together, a retinal image for calculating a focus value is generated by photographing the retina while moving the focus. The focal interval may be associated with a time interval between points at which the second camera 232 captures retinal images. Here, the focal intervals may be equal intervals or variable intervals, and may be expressed as the rotation number and/or rotation angle of the motor rotation shaft.

Equal interval auto focusing calculates focus values at the same focus interval from an initial position to a final position. When the rotation axis of the motor rotates at a constant speed and the focus position changes at a constant speed, the second camera 232 photographs the retina at the same time interval, and a focus value for the generated retinal image may be calculated. Since the focus is moved from the initial position to the final position, if the focus interval is set sufficiently small, the maximum focus value can be accurately determined, but the calculation time may increase.

The variable interval auto focusing may reduce a maximum focus value calculation time. The variable-interval auto-focusing calculates a focus value while reducing a focus period for obtaining a retinal image. There are two or more focal sections, and at least a part of each focal section may overlap. Each focal interval may have a different focal interval. Meanwhile, the rotation shaft of the motor may rotate at a constant speed or at a different speed in all focal sections.

In one embodiment, there are three focal sections, and a focal interval may be different for each focal section. Auto focus moves the focus from the initial position R ₀ at a first speed in the first focal interval, and the second camera 232 is located at positions R ₁ , R ₂ , R ₃ , spaced apart by a first focal interval. A retinal image is captured in R ₄ (S610). The first focal section may be substantially the same as an entire focal section defined by an initial position and a final position in equal interval auto focusing. Meanwhile, the first focal interval and the first speed may be greater than the focal interval and speed in equal interval auto focusing.

Focus values related to the positions R ₁ , R ₂ , R ₃ , and R ₄ are calculated, and a second focus section is determined from the calculated focus values (S610). The second focal region may be defined as two contiguous positions selected from positions R ₁ , R ₂ , R ₃ , and R ₄ . In the graph illustrated in FIG. 25 , the focus value at position R ₂ is the largest. Among positions R ₁ and R ₃ subsequent to position R ₂ , the focus value at position R ₃ is greater than the focus value at position R ₁ . Accordingly, the initial position of the second focal region is set to position R ₂ and the final position is set to R ₃ .

Auto focus moves the focus from the initial position R ₂ at a second speed in the second focal interval, and the second camera 232 is located at positions r ₁ , r ₂ , r ₃ , spaced apart by a second focal interval. A retinal image is captured in r ₄ (S630). The second focal interval is smaller than the first focal interval, and the second speed may be equal to or smaller than the first speed.

Focus values related to the positions r ₁ , r ₂ , r ₃ , and r ₄ are calculated, and a third focus section is determined from the calculated focus values (S640). The third focal region may be defined as two contiguous positions selected from positions r ₁ , r ₂ , r ₃ , and r ₄ . In the graph illustrated in FIG. 25 , the focus value at position r ₂ is the largest. Among positions r ₁ and r ₃ consecutive to position r ₂ , the focus value at position r ₁ is greater than the focus value at position r ₃ . Accordingly, the initial position of the third focal region is set to position r ₁ and the final position is set to r ₂ .

The auto focus moves the focus from the initial position r ₁ at a third speed in the third focal interval, and the second camera 232 captures a retinal image at positions spaced apart by a third focal interval (S640). . The third focal interval may be smaller than the first focal interval and the second focal interval, and the third speed may be equal to or smaller than the second speed.

A maximum focus value is determined among the focus values calculated in the third focus section (S650). Auto focus moves the focus to a focus position related to the maximum focus value, and then retinal imaging ( S40 ) is performed.

20 is a functional view of a binocular imaging device.

The binocular imaging devices 100 and 101 include left eye imaging units 110L and 111L and right eye imaging units 110R and 111R tilted at a predetermined angle toward both eyes, and a controller 600 controlling them. The controller 600 may be implemented by a program running on a physical device such as a processor, memory, I/O, and the like. The controller 600 may include a first camera controller 610, a second camera controller 620, an image processor 630, a lighting controller 640, and a movement mechanism controller 650.

The first camera controller 610 controls the first camera 210 to capture a pupil to generate a first tracking image, and the second camera controller 620 controls the second camera 232 to capture a specific target on the retina. is photographed and controlled to generate a second tracking image and a retinal image. The first camera controller 610 and the second camera controller 620 transmit camera driving signals necessary for camera driving, such as focus and exposure time, to the first camera 210 and the second camera 232 . Meanwhile, the first camera controller 610 and the second camera controller 620 may transmit to the lighting controller 540 a lighting trigger signal requesting turning on/off of lighting necessary for tracking and photographing operations.

The image processing unit 630 processes the first tracking image and/or the second tracking image to generate a virtual image for inducing pupil movement. Also, the image processing unit 630 may transmit a movement trigger signal for driving the movement mechanisms 130 and 140 to the movement mechanism controller 650 according to the processing result of the first tracking image.

The lighting controller 640 generates a lighting control signal for turning on/off lighting required for each operation. The first light 200 , the second light 400 , and the white light 220 are turned on/off by the light control signal.

The movement mechanism control unit 650 controls the movement mechanisms 130L, 130R, and 140 to move the left eye capture units 110L and 111L and the right eye capture units 110R and 111R in a first direction and/or a second direction. do.

The above description of the present invention is for illustrative purposes, and those skilled in the art can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. In particular, the features of the present invention described with reference to the drawings are not limited to the structures shown in the specific drawings, and may be implemented independently or in combination with other features.

The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts thereof should be construed as being included in the scope of the present invention. .

Claims (8)

a left eye photographing unit having a first optical axis twisted at a first angle with respect to the optical axis of the left eye and capturing a pupil and retina of the left eye;
a right eye photographing unit having a second optical axis twisted at a second angle with respect to the optical axis of the right eye, and capturing a pupil and retina of the right eye;
a first direction movement mechanism for moving the left eye capture unit and the right eye capture unit in a first direction to align an optical axis of the left eye capture unit with the pupil of the left eye and align an optical axis of the right eye capture unit with the pupil of the right eye; and
A second direction movement mechanism for moving the left eye capture unit and the right eye capture unit in a second direction toward both eyes of the subject,
The left eye capture unit and the right eye capture unit are arranged to converge toward both eyes of the examinee,
The left eye capture unit and the right eye capture unit,
a retinal illumination system for respectively incident ring-shaped white polarized light to the left eye and the right eye along the first optical axis;
a pupil tracking optical system that is optically coupled to the retinal illumination system and generates a first tracking image by photographing pupils of the left eye and the right eye; and
A retina tracking/photography optical system that is optically coupled to the retinal illumination system, photographs the retina to generate a second tracking image, and generates a retinal image by detecting white polarized light reflected from the retinas of the left eye and the right eye, respectively. and
The second tracking image is blurred and used to track the position of the optical disc in the retina. The binocular imaging device of claim 1 , wherein the left eye capture unit and the right eye capture unit simultaneously capture the left eye and the right eye, respectively, to generate a retinal image. delete delete The binocular imaging device of claim 1 , wherein a focal point for imaging the optical disk in the retina is determined by a focus value that is a sum of pixel value differences between pixels belonging to an optical disk corresponding region tracked by the left eye photographing unit. . The binocular imaging apparatus of claim 5 , wherein the focus value is calculated after blurring the area corresponding to the optical disk. The binocular imaging apparatus of claim 5 , wherein the focal position is determined by a maximum value among focus values calculated from a plurality of optical disk corresponding regions obtained by photographing at a constant focal interval in one focal period. The binocular imaging of claim 5 , wherein the focal position is determined by a maximum value among focus values calculated from a plurality of optical disk corresponding regions acquired by photographing at different focal intervals in a plurality of focal sections in which some sections overlap. Device.

Images