JPH025A - Device for detecting eye direction of camera - Google Patents

Abstract

PURPOSE:To accomplish the detection of a photographer's eye direction by projecting an infrared ray on an eye, catching a reflected light from a firs Purkinje image based on the mirror reflection of a cornea and the reflected light from an eyeground, and arithmetically operating a photoreceiving output. CONSTITUTION:The infrared ray is projected from an infrared light source 48 on the photographer's eye which is positioned on the right side of the pentaprism 40 of the eye direction detecting device 46. Thereby, the first Purkinje image PI based on the mirror reflection of a cornea is formed. The reflected light from the Purkinje image and the reflected light from the eyeground are caught by means of a primary line sensor 53. The light receiving output of the sensor 53 is amplified by an amplifier 56 and a digital signal is converted by an analog digital converter 57. Then, the process of detecting the photographer's eye direction is performed by means of a microcomputer 58. Consequently, the eye direction can be detected, and the optical systems of plural focusing zones can be automatically selected and driven.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a line-of-sight direction detection device for a camera, and more particularly, to a device that is optically approximately conjugate with a plurality of focusing zones provided within the field of view of a finder. A focusing zone of the autofocus optical system corresponding to each focusing zone of the finder is set at a position, and each focusing zone of the finder is selected and the selected focusing zone is set. The present invention relates to a line-of-sight direction detection device suitable for a camera having an automatic focus device that uses an autofocus optical system corresponding to the zone to focus on a subject that appears to overlap a focusing zone.

(Background of the Invention) Conventionally, some cameras are equipped with an autofocus optical system. For example, Fig. 39 shows a schematic configuration of the optical system of a single-lens reflex camera equipped with this autofocus optical system. 5 is a condenser lens, 5 is an aperture mask, 6 and 7 are separator lenses, and 8 is a CCO as a light receiving section. Here, field mask 3.
Condenser lens 4. Aperture mask 5. Separator lens 6? , CCD8 are integrated into a module.

An autofocus optical system 9 is configured.

In this autofocus optical system 9, the field mask 3 is provided near the film equivalent surface 10.

The film equivalent surface 1G is exposed to the subject 2 through the photographic lens 1.
It is in an optically conjugate position with. This film equivalent surface 10
When the photographic lens 1 is in focus, the subject 2
An image 11 of is formed in focus. The condenser lens 4 and the aperture mask 5 have the function of splitting the photographing light passing through the left and right sides of the photographic lens l into two luminous fluxes, and the separator lens 6.7 is connected to the photographic lens 1 via the condenser lens 4. is in a conjugate position.

The separator lens 6.7 is arranged horizontally, as schematically shown in FIG. This separator lens 6.7 is connected to a virtual aperture area 14.1 of the exit pupil 13 of the photographic lens 1 via a focusing zone 12 located at a position optically conjugate with the central focusing zone of the finder, which will be described later.
I'm looking at 5. The separator lens 6.7 receives the light beam that has passed through its aperture area 14.15. The image ti formed on the film equivalent surface 10 is divided into two areas of the CCD 8 by the separator lens 6.7.
The image is re-imaged as

The interval of the signal S corresponding to the image interval when the re-focused image 11 is in focus (see FIG. 41(a)) is expressed as shown in FIG. shall be. Here, when the photographing lens 1 is in focus on the m side, the image interval becomes narrower as shown in FIG. 42, compared to when the camera is in focus, as shown in FIG. 41(b). The interval between signal S is sad. becomes smaller than On the other hand, as shown in FIG. 41(c), when the photographic lens 1 is in focus at the rear side compared to when it is in focus,
As shown in FIG. 42, the image interval has widened and the corresponding interval of the signal S has become narrower. becomes larger than This change in the image interval is proportional to the defocus amount of the photographing lens 1, so in a conventional single-lens reflex camera, the image interval of the CCD 8 is detected and this is processed to determine the defocus direction of the photographic lens 1. Depending on the amount of defocus.

The photographic lens 1 is driven to the in-focus position.

For example, as shown in FIG. 43, when composing the photograph so that the desired subject 2 falls within the central focusing zone 17 provided at the center of the finder 16, and operating a button, the defocus direction and defocus direction can be adjusted. The amount is automatically calculated, and a photograph with the subject 2 in focus can be obtained.

By the way, in this type of single-lens reflex camera, the focusing zone is provided at the center of the finder 16, so if this is done, the subject 2 will be located at the center of the photograph.

However, there may be cases where it is desired to obtain a photograph in which the subject 2 is placed in the vicinity.

In consideration of this, conventional single-lens reflex cameras are provided with a focus lock mechanism. If you use this focus lock mechanism number, you can position the subject 2 in the center of the viewfinder, focus the photographic lens on the subject 2, lock the focus in this state, and frame and photograph as shown in No. 441i1. , it is possible to obtain a photograph in which the desired subject 2 is placed in the periphery.

However, with this single-lens reflex camera, you have to recompose the shot before taking the picture, which can sometimes be too time-consuming.

Therefore, the present applicant previously filed an application for an automatic distance measuring device for a single-lens reflex camera that can quickly perform photographing operations to obtain photographs with a desired subject located in the peripheral area (
Special [@62-22561].

The device disclosed in the application will now be briefly explained with reference to FIGS. 27 to 30.

In FIG. 27, 13 indicated by a solid line is an exit pupil seen from the focusing zone 12 of the autofocus optical system 9.

This exit pupil 13 is approximately circular as shown in FIG.

° On the other hand, the aperture area 1 seen from the separator lens 6.7
4.15 is approximately elliptical.

Autofocus optical systems I11.19 for peripheral focusing are provided on both the left and right sides of the autofocus optical system 9. The autofocus optical system 18 has a pair of separator lenses 20, 21-CCD 22, and the autofocus optical system 19 has a pair of separator lenses 23, 24,
It has CCD25.

As shown in FIG. 29, within the field of view of the finder 16, autofocus optical systems 1g and 19 for peripheral focusing are provided on both sides of the central focusing zone 17.
Peripheral focusing zones 26, 27 are arranged side by side.

The peripheral focusing zones 26 and 27 are in an optically substantially conjugate positional relationship with the autofocus focusing zones 2g and 29. Separator lens 20.21. The separator lenses 23 and 24 are arranged in the vertical direction, and are optically substantially conjugate with the exit pupil 13 shown by the broken line of the photographing lens 1 via the condenser lens 4 (not shown), and are located in focusing zones 28 and 29.
The user is looking into the vertical aperture region 30.31 of the exit pupil 13 indicated by the broken line.

In this way, the separator lenses 20.21. The reason why the separator lenses 23 and 24 are arranged in the vertical direction is that the light flux that enters the focusing zone 28, 29 via the photographing lens l is
As shown in No. 30111, the beam becomes oblique due to the influence of vignetting, and the exit pupil 13 of the photographing lens 1 shown by the broken line as seen from the focusing zone 28, 29 is vignetted and has a flattened shape. , when the opening area 30.31 is provided in the horizontal direction, the separator lens 20.21 (
A sufficient baseline length between the lenses of the separator lenses 20 and 21) cannot be secured.

This is because the performance of the lens deteriorates and the accuracy of detecting the image interval deteriorates.

In FIG. 27, breath is the optical axis of the photographing lens 1, □ is the central optical axis of the autofocus optical system 18, and jl
, are the central optical axes of the autofocus optical system 19, and the central optical axes 1 and 2 intersect at the center 01 of the exit pupil 13, which is indicated by a solid line. Also.處, 1 is the optical axis of the separator lens 20, □, is the optical axis of the separator lens 21, n2 □ is the optical axis of the separator lens 23, att is the optical axis of the separator lens 24, ．． □ is opening area 3
It intersects at the center 02 of 1'.

The optical axes Q□7+a,2 intersect at the center 03 of the aperture area.

In this way, a plurality of focusing zones are provided within the field of view of the finder 16, and an autofocusing zone corresponding to each focusing zone of the finder 16 is provided at a position that is optically substantially conjugate with the plurality of focusing zones. A focusing zone is provided for the focusing optical system, and in order to drive the CCD corresponding to the focusing zone intended by the J11 shadow person (see Figure 29), the intended focusing zone is selected by button operation. By doing so, the photographing lens can be automatically focused on the subject 2 that can be seen through the selected focusing zone using the autofocus optical system corresponding to the selected focusing zone. Become.

Therefore, using this single-lens reflex camera eliminates the tendency to lock the focus to determine the composition.

(Problem to be Solved by the Invention) By the way, it is possible to provide a plurality of focusing zones 17, 26, 27 within the field of view of the finder 16, and to Each focusing zone 1 of the finder 16 is located at an optically approximately conjugate position.
Since focusing zones 9, 18, and 19 of the autofocus optical system corresponding to 7, 26, and 27 are provided, any one of the plurality of focusing zones within the field of view of the finder 16 is selected. By making it possible to automatically detect whether or not the object is in focus, the trouble of manually selecting one of the plurality of focusing zones 17, 26, 27 provided within the field of view of the finder 16 can be eliminated. This makes the camera even more convenient.

The present invention has been made in view of the above circumstances.

A first object of the present invention is to provide a camera line-of-sight direction detection device that detects the line-of-sight direction of a photographer's eyes.

A second object of the present invention is to align an autofocus optical system corresponding to each focusing zone of the finder at a position that is optically approximately conjugate with a plurality of focusing zones provided within the field of view of the finder. Focusing zones are provided, any one of the focusing zones of the finder is selected, and an autofocus optical system corresponding to the selected focusing zone is used.

It is an object of the present invention to provide a line-of-sight direction detection device for a camera suitable for a camera having an automatic focusing device that focuses on a subject that appears to overlap the focusing zone.

A third object of the present invention is to provide a camera line-of-sight direction detection device that detects the line-of-sight direction of a photographer's eye using a one-dimensional line sensor.

(Means for Solving the Problems) The camera line-of-sight detection device according to the present invention is characterized by having a light transmitting system that guides a parallel light beam to the photographer's eye, and a light receiving section, and based on corneal specular reflection of the eye. a light receiving system that receives reflected light forming a first Purkinje image and reflected light from the fundus of the eye; and a processing circuit that detects the line of sight direction of the photographer's eye based on the light receiving output of the light receiving section. It's located on the camera body.

Another feature of the camera line-of-sight detection device according to the present invention is that a plurality of focusing zones are provided within the field of view of the finder of the camera body, and the focusing zone is located at a position substantially optically conjugate with the focusing zone. The focusing zone of the autofocus optical system corresponding to the focusing zone is provided, and the processing circuit is configured to automatically sense that any one of the focusing zones of the finder is selected. be.

Further features of the camera line of sight detection device according to the present invention include:
The light receiving section is composed of a one-dimensional line sensor, and the processing circuit converts the output from the one-dimensional line sensor into the fundus reflected light corresponding output degree component corresponding to the reflected light from the fundus and the reflected light forming the first Purkinje image. and a separation means for separating the output component corresponding to the first Purkinje image forming reflected light, and determining the center of gravity position of the separated output component corresponding to the fundus reflected light and the center position of the center of gravity of the output component corresponding to the first Purkinje image forming reflected light, respectively. , which detects the direction of the eye's line of sight.

Other features will become apparent from the description of the invention.

(Principle of the invention) First, before explaining embodiments, the principle of the present invention will be explained.

A detection method for detecting the direction of the line of sight is described, for example, in ``Psychophysics of Vision'' by Mitsuo Ikeda, but when applied to a camera, it must be done so that parallel movement of the photographer's eye is not detected. , because in the case of detecting the direction of the line of sight of the eye as well as the parallel movement of the eye,
This is because the information on the line of sight direction due to the parallel movement of the eyes overlaps the information on the angular direction, making it impossible to distinguish which focusing zone the photographer is gazing at.

If we were to use a line-of-sight direction detection optical system that can detect parallel movement, we would have to keep the relative distance between the optical axis of the camera's viewfinder and the center of rotation of the photographer's eyeball constant; Considering that hand-held cameras are common, this is virtually impossible because the eye moves left and right relative to the finder 16.

As a line-of-sight direction detection optical system that detects line-of-sight only in the angular direction, for example, the Optical fin in 1974
Gineering magazine 77 August issue VOL, 13. NO
4, Fixation Poi on P339-P342
nt Naasuraw+ent by theOcu
There is one introduced in lometar Technique.

The principle of the line-of-sight direction detection optical system advocated in this article is that, as shown in FIG. The image of the light source located at is generated as a light point at a midpoint Q between a light point K where the center of curvature k of the convex mirror 30 and the optical axis 3 intersect with the mirror surface.

Here, even when the cornea 32 of the human eye 31 is irradiated with a parallel light beam P parallel to the optical axis as shown in FIG. Center of curvature R
and the corneal vertex K' as a light spot (this light spot is called the first Purkinje image PI), and the symbol 3
3 is the iris, 34 is the center of the pupil, and SA' is the center of rotation of the eyeball.

When the optical axis 1 of the luminous flux P illuminating the corner 132 coincides with the visual axis 1, which indicates the line of sight direction of the human eyes.

Center of pupil 34. First Purkinje statue P1. The center of curvature R of the cornea 32 and the center of rotation SA' of the sphere are above the optical axis ml. When thinking about a camera, the optical axis of the finder is 1. Assuming that the center of rotation SA' of the eyeball is located above.

Assume that the eyeball is rotated in the left-right direction around the center of rotation SA'. Then, as shown in FIG. 24, the center 34 of the pupil
A relative shift occurs between the first Purkinje image PI and the first Purkinje image PI.

Also. Suppose that the eye is rotated by an angle of 0 with respect to the optical axis 11, and the length of the perpendicular line drawn from the center 34 of the pupil to the light ray P' that perpendicularly enters the cornea 32 is d.

d=k Yu・sin II ・・・・・・■Here, k
□ is the distance from the center 34 of the pupil to the center of curvature R of the cornea 32, and although there are individual differences, it is IL-HDBK-141rOPTICAL as edited by the United States Department of Defense.
According to DESINJ, it is about 4.5cm. Note that the symbol H indicates the intersection of the light ray P' and a perpendicular line drawn from the center 34 of the pupil to the light ray P' that perpendicularly enters the cornea 32.

As is clear from the above equation (2), since the distance k is known, the rotation angle 0 can be found by finding the length d.

Here, considering that the intersection point H and the first Purkinje image PI are on the ray P', the parallel light beam P is irradiated toward the cornea 32, and among the specularly reflected light from the cornea 32, Ray P that is reflected and returned in a direction parallel to the incident light beam
” and find the relationship between the center 34 of the pupil and the first Purkinje image PI, the rotation angle 0 of the eye can be determined.

Therefore, the parallel light beam P is projected onto the eye, and the 25th! l-26th
As shown in the figure, when the periphery 34' of the pupil that stands out as a silhouette based on the light reflected from the fundus and the first Purkinje image PI are imaged on a light receiving element (for example, a one-dimensional line sensor), the light received The received light output on the element has a peak at a location corresponding to the first Purkinje image PI, and the location corresponding to the reflected light from the fundus has a trapezoidal shape. ,34
The pupil periphery corresponding coordinates 11+1. At the same time, the coordinates corresponding to the first Purkinje image PI corresponding to the first Purkinje image PI are obtained using the slice level 2, and the center coordinate i' corresponding to the center 34 of the pupil is obtained using the following formula and Chuoza I! Calculate the difference d=PI-i from IPI. Here, if the magnification of the detection optical system is m, then the distance d can be obtained from the following equation.

i == (in+ i 1) / 2...
・■p■'= (pzl+pu,)/2 ・・・・
...0d=d/m ...■ Therefore, if a line-of-sight direction detection device equipped with such a processing circuit is used, it is possible to detect which of the plurality of focusing zones provided in the finder 16. This means that you can automatically choose whether or not you are looking at nine.

(Example) Hereinafter, an example of the line-of-sight direction detection device for a camera according to the present invention will be described with reference to the drawings.

In FIG. 1, 40 is a pentaprism built into the camera, and 41 is a quick return mirror.

42 is a focusing plate, 43 is a condenser lens, 44 is a finder magnifying glass, 45 is the photographer's eye, and is the optical axis of the finder optical system described in I18ft4. Here, the finder magnifying glass 44 is a device camera body composed of lenses 44a and 44b, and has a pentaprism 40 on the opposite side of the finder magnifying glass to detect the line of sight direction of the photographer's eye 45 looking into the finder l6. Gaze direction detection device! i46 is included. FIG. 1 shows a frame 47 of the line-of-sight direction detection device 46. Gaze direction detection device [46
has a light transmitting system 46A and a light receiving system 46B. Light transmission system 46
As shown in FIGS. 2 and 3, A has an infrared light source (for example, an infrared light emitting diode) 48 that generates infrared light. This infrared light is transmitted through a half mirror 49, a reduction lens 50,
The light is irradiated to the photographer's eye 45 as a parallel beam of light via the Fumbenseta prism 51, the pentaprism 40, and the finder loupe 44. As a result, a first Purkinje image PI based on specular reflection of the cornea 32 is formed.

The reason why infrared light was used here was to avoid giving the photographer glare caused by the illumination of the optical system of the line-of-sight direction detection device [46]. On the other hand, the reason why we decided to use the reduction lens 50 is as follows.

first. This is because the optical path length of the optical system of the line-of-sight direction detection device [46] was made as short as possible so that it could be compactly incorporated into the camera. Next, optical axis j1. Since only the infrared reflected light parallel to the rays is used, the amount of reflected light from the eye 45 is thought to be small. This is because consideration was given to increasing the sensitivity on the light-receiving surface of the element.

Of the light reflected from the cornea 32 of the eye 45, the light beam parallel to the incident light beam is transmitted to the finder loupe 44, the pentaprism 40, the compensator prism 51. reduction lens 5G
is guided to the half mirror 49 through the
49 to the reimaging lens 52.

The image forming lens 52 forms an image on a one-dimensional line sensor (for example, CCD) 53 as a light receiving element by the re-imaging lens 52. As shown in FIG.
4, the mask 54 is provided with an opening 55, and the center of the opening 55 is the center of curvature Y of the re-imaging lens 52.
Located in Here, the diameter of the -a port 55 is approximately L2 mm.

The photographer's eye 45 is normally placed at the eye point, and the one-dimensional line sensor 53 and the photographer's eye 45 are connected to each other.
The pupil is shown schematically in Figure 5. They are assumed to be in an optically conjugate positional relationship via the finder magnifying glass 44, reduction lens 50, and reimaging lens 52.
On the one-dimensional line sensor 53, together with the first Purkinje image PI, the periphery 34' of the pupil is formed as a silhouette by the reflected light from the fundus. Therefore, as shown in FIG.
The signal is amplified by the analog-to-digital converter 57, converted into a digital signal, and temporarily stored in the memory 59 of the microcomputer 58.

The distance k1 is recorded in the memory 59 as information. The information on this distance k1 and the information on the light receiving output are called into the calculation circuit 60, and calculations are performed based on the formulas 1 to 2 to obtain the rotation angle θ, and which focusing zone is selected from this rotation angle θ. The drive circuit 61 is caused to output selection number 44, which means whether the selection is made or not.

When the drive circuit 61 drives the CCD of the autofocus optical system corresponding to the selected focusing zone, the photographing lens is automatically focused on the subject that can be seen through the focusing zone intended by the photographer. can be done.

By the way, as shown in FIG. 29, left and right focusing zones 0, . The distance to 0 (image height) is y
Assume that the focal length of the finder magnifying glass 44 is f.

y=f+tan　θ　・・・−・・・■ Substituting the ■expression into the above ■expression.

F=f−d/(K,·eos 6) −−■ That is, y is proportional to d/(K,·cost).

This means that even if the distortion of the image formed on the one-dimensional line sensor 53 is eliminated, the value of y cannot be determined linearly from the value of d.

It means the existence of nonlinearity.

35 ■■ For vignetting, etc. in the case of a camera.

The image height y of the plurality of focusing zones is at most 6 -
~9-臘.

Here, it is assumed that the optical system of the line-of-sight direction detection device 46 transmits the pupil image with nonlinearity to the rear one-dimensional line sensor 53, and that the one-dimensional line sensor 53
Assuming that the length d detected in is proportional to the image height y, the length is 0.1% to 1.6% smaller than the actual length d.
Although only the longer gourd is detected, there is no problem in selecting the focusing zone, but from the viewpoint of improving the accuracy of the optical system of the line-of-sight direction detecting device i46, it is preferable that there be no nonlinearity.

In such a case, correction can be made using a microcomputer. However, if distortion exists in the optical system itself, measurement will be inaccurate, so it is necessary to eliminate at least distortion in the optical system.

Therefore, in order to reduce the spherical aberration of the reduction lens 500, the surface 50a on the side closer to the finder magnifying glass 44 is made an aspherical surface, and the focal point of the reduction lens 50 is positioned at the center of curvature Y of the re-imaging lens 52. In this way, the reduction lens 50
When is made an aspherical surface and the focal point of the reduction lens 50 is located at the center of curvature Y of the re-imaging lens 5z, the aperture 55 is located at the center of curvature Y of the re-imaging lens 52, resulting in less distortion. This makes it possible to realize an optical system, which is even more preferable as an optical system for the line-of-sight direction detection device [46].

Next, an example of the design of the optical system of such a line-of-sight direction detection device 46 will be described below.

First, set the distance from the lens 44a to the eye point to 14
．． 7l-, the center thickness of the lens 44a is 4.98 mm,
The radius of curvature of the eye point side surface of the lens 44a is convex 1
gt, tea-1, the radius of curvature of the surface of lens 44a facing lens 44b is -25,500■■, lens 4
The refractive index of 4a is 1.69105. The distance between the lenses 44a and 44b at the top of the optical axis is 3. Ol
■Serve as a meal. Also. The center thickness of the lens 44b is 4.10
The radius of curvature of the surface of the lens 446 facing the lens 44a is concave -23,860m+m, and the radius of curvature of the surface of the lens 44b facing the pentaprism 40 is convex -48,
140 mm, and the refractive index of the lens 44b is 1.79175. In addition, the surface 40a of the pentaprism 40 and the lens 4
The distance from 4b is 3.21mm, and the pentaprism is 40mm.
The length from the zero surface 40a to the surface 40b along the optical axis is 28.00 mm. The radius of curvature of each surface 40a, 40b is .

The refractive index of the pentaprism 40 is 1.51260.

Next, the distance between the surface 51a of the compensator prism 51 and the surface 40b of the pentaprism 40 is set to 0.10 mm, and the distance between the surface 51b of the compensator prism 51 and the surface 50a of the reduction lens 50 is also set to 0.10 mm. The surface 51b and surface 5 of the gompensator prism 51
The length at the top of the optical axis with 1a is + 2.00 mm,
Each side 51a.

The radius of curvature of 51b is φ, and its compensator prism 5
The refractive index of 1 is 1.51260.

The reduction lens 50 has a convex radius of curvature of 12.69 on the surface 5◎a.
0m garden (k3=-3,00), its center thickness is designed to be 2.00su+, and its refractive index is 1.4J.
1716. In addition.

The radius of curvature of the other surface 50b of the reduction lens 50 is convex -2
00,000 mm, and the reimaging lens 52 and its surface 5
The interval with 0b is set to 11.48 degrees.

The radius of curvature of the surface 52a of the re-imaging lens 52 is convex with a radius of 1.52
0 ■ The radius of curvature of the surface 52b is Φ, and the re-imaging lens 5
The center thickness of lens 2 is 1.52 mm, and the refractive index is 1.48716, which is the same as that of the reduction lens 50. Diameter 0.2
Since the mask 54 having an opening 55 of mm is attached to the surface 52b, the distance between the mask 54 and the surface 52b is Ol.
The thickness of the mask 54 is 0.04 mm, and the distance from the mask 54 to the light receiving surface of the light receiving element 53 is 1.46 mm.
mm. Note that the radii of curvature of the light-receiving surfaces of the mask 54 and the light-receiving element 53 are such that air is present between each optical element.

Further, +k3 indicates an aspheric coefficient, and there is a relationship between it and the sag amount X as shown in the following equation.

X=h-c/(1+J l-(k3+IJ h-c-J
Here, h indicates the height from the optical axis a, and C is the reciprocal of the radius of curvature of the reduction lens 50.

If the reduction lens 50 is not made of an aspherical surface, spherical aberration will occur as shown in FIG. 6, and distortion as shown in FIG. 7 will occur, but the line-of-sight direction detection optical system designed as described above is used. As shown in FIG. 8, spherical aberration is improved, and along with this, distortion is improved as shown in FIG. 9.

In this embodiment, an LED corresponding to each focusing zone 17, 26, 27 is provided within the field of view of the finder 16, and an LED corresponding to a selected focusing zone is provided.
It is also possible to have a configuration in which D is displayed blinking to confirm whether or not the focusing zone is the one intended by the photographer. Also,
In this embodiment, there are three
Although the case where there are two or more focusing zones has been described, the case where there are two or more focusing zones is applicable.

It will be easy to understand that the present invention is realized.

Furthermore, in this embodiment, the light transmitting system 46A and the light receiving system 46B are installed on the opposite side of the finder magnifying glass 44 with the pentaprism 40 as the boundary.

Either one of the light transmitting system 46A and the light receiving system 46B may be provided on the same side of the pentaprism 40 as the finder magnifying glass 44. This will be discussed later.

Next, another embodiment of the line-of-sight direction detecting device 46 according to the present invention will be described with reference to FIGS. 10 to 13.

A two-dimensional solid-state image sensor can also be used for the light receiving section. However, in this case, since the array of solid-state image sensors is two-dimensional, it is expected that the scanning processing time for scanning the solid-state image sensors will take a long time.

Moreover, the cost is also high. By the way, the center O,,0,,0, of the plurality of focusing zones 17, 26, 27 is the 29th
For things arranged in a straight line as shown in @.

The center of the focusing zone 17.26.27 O,, O,,
It is conceivable to use a one-dimensional line sensor in which photoelectric elements are arranged in a direction corresponding to the direction in which 0 and 0 are lined up. However, when such a one-dimensional line sensor is used, there are problems as described below.

FIGS. 12 and 13 are diagrams for explaining this problem. In FIG. 12, lOO is a finder magnifying glass, 101 is a re-imaging lens, and 102 is a one-dimensional line sensor. As shown in this figure, the line of sight direction detection device! 46
The optical axis of the optical system! ,,i.e., Finder Loupe 10G
The optical axis emperor, and the visual axis of the human eye 31! , ' match, the pupil image 34 as the silhouette (periphery) of the pupil
a, the first Purkinje image PI is the one-dimensional line sensor 10
2, the line of sight direction can be detected normally. However.

When the human eye 31 moves vertically with respect to the camera body, the pupil image 3 becomes a silhouette as shown in FIG.
4a, the first Purkinje image PI is the one-dimensional line sensor 10
2, causing the inconvenience that line-of-sight direction cannot be detected normally.

Therefore, as shown in FIG. 11, the re-imaging lens 52
For example, a silicon trical lens is used.

A mask 54 having the same structure as shown in FIG. 4 is provided on the flat surface side of this silicon trical lens. The mask 54 is provided with an aperture 55, and the center of the aperture 55 is located at the center of curvature Y of the re-imaging lens 52. Here, the opening 55 is a rectangular slit hole, and the extending direction of the slit hole is orthogonal to the arrangement direction of the photoelectric elements 53a of the one-dimensional line sensor 53. The re-imaging lens 52 has its curved surface disposed on the finder magnifying glass 44 side.

, thus, the photoelectric element 53 of the one-dimensional line sensor 53
11 are arranged corresponding to the focusing zones of a plurality of autofocus optical systems, a cylindrical lens is used for this re-imaging lens 52, and the re-imaging lens 52 is arranged perpendicularly to the arrangement direction of the one-dimensional line sensor 53. vertically in the direction of
Since the arrangement is such that the Purkinje image PI and the pupil image 34a as a silhouette are formed on a plane including the one-dimensional line sensor 53, the eye 45 is vertically aligned with respect to the camera body, as shown in FIG. Each of those images PI.

At least a part of 34a is formed in the one-dimensional line sensor 102. In addition, the opening 5 of the mask 54
5 is also a slit hole that extends long in the direction perpendicular to the arrangement direction of the photoelectric elements 53a of the one-dimensional line sensor 53.
A pupil image 34 formed on a surface including the one-dimensional line sensor 53. The first Purkinje image PI becomes more vertically elongated in the direction perpendicular to the arrangement direction, so that the line-of-sight direction can be detected reliably.

In this embodiment, a cylindrical lens is used as the re-imaging lens 52, but a toric lens may also be used.

Next, another example of the processing circuit of the line-of-sight direction detection device 46 according to the present invention will be described.

In order to incorporate the optical system of the line-of-sight direction detection device 46 into the camera body and to avoid cost increases as much as possible, it is desirable that the optical system be as simple as possible. It is preferable that there be.

However, when such a re-imaging lens 52 is used,
When light with a uniform light intensity distribution is made incident on the re-imaging lens 52, as schematically shown in No. 1414.

The amount of light that is imaged on the light receiving surface of the one-dimensional line sensor 53 is attenuated at the periphery. In that Figure 14.

The two-dot chain line G shows the light amount distribution in the case that there is no light amount attenuation, the broken line G shows the light amount distribution in the case that there is light amount attenuation, and , as in the above, indicates the optical power of the line-of-sight direction detection device 46. It shows the optical axis of the system.

If the center of gravity of the light intensity distribution is determined based on the output of the one-dimensional line sensor 53 under such light intensity attenuation, there is a risk that the obtained center of gravity will deviate from the actual center of gravity. If the line-of-sight direction is determined by calculation using , an error will occur between the line-of-sight direction and the actual line-of-sight direction.

If the angles of the line-of-sight directions to be distinguished are far apart, this error based on light attenuation can be tolerated, but as the angle of line-of-sight directions to be distinguished becomes smaller, the error based on light attenuation cannot be ignored. . The present invention is not limited to this, but as long as the error due to light intensity attenuation can be removed, it is preferable to remove it as much as possible in order to detect the line-of-sight direction by arithmetic processing.

Therefore, in this processing circuit, a method is taken in which the light amount attenuation is determined in advance and a light amount correction value is stored in a ROM which will be described later.

That is, the output distribution of the one-dimensional line sensor 53 corresponding to a light amount distribution with light amount attenuation is as indicated by the symbol G3 in the 1411th line. Here, the code i means the i-th photoelectric element 53a, j means the j-th photoelectric element 53a, X is the output of the 111th photoelectric element 53a, and Xj is the output of the j-th photoelectric element 53a. It shows. Now, it is assumed that the j-th photoelectric element 53a is located above the optical axis. That is, this j-th photoelectric element 53a has addresses a and b.
Assume that the address is in the middle of the street address. In this case, the output of the j-th photoelectric element 53a can be expected to be maximum.

Therefore, the force of Fuyuyama from the photoelectric element 53a at address a to the photoelectric element 53a at address b is determined, and the correction coefficient H is determined.

The following relational expression exists between the correction coefficient H, the output X, and the output Xj.

H,--X, = X, -@And, in order to normalize this correction coefficient H, divide it by xl to get the correction value HL
' is determined and stored in the ROM of the processing circuit shown in FIG.

HL = HL /
By multiplying and correcting the output of the photoelectric element 53a (from address a to address b), the output distribution corresponding to the attenuated light amount distribution is corrected, as shown by symbol G. That is, for uniform light.

A uniform output distribution G is obtained in which the attenuation of the amount of light due to the influence of the peripheral portion of the re-imaging lens 52 is corrected.

Further, as a correction value, a correction value based on the light amount distribution obtained when parallel uniform light is incident from the finder magnifying glass 44 is used, and this is written into the rewritable El! If stored in PRON, errors based on the light intensity distribution including optical elements of the optical system other than the re-imaging lens 52, and variations in sensitivity of each photoelectric element 53a of the one-dimensional line sensor 53 itself can be included. Corrections can be made later. Therefore, by performing such correction, it becomes possible to relax the standards regarding the optical characteristics of the one-dimensional line sensor 53 itself, and it is possible to reduce costs based on improved yield.

By the way, the first Purkinje image PI based on corneal specular reflection
In order to obtain the center of gravity of the light intensity distribution that forms the fundus and the center of gravity of the light intensity distribution of the light reflected from the fundus, the output of the one-dimensional line sensor 53 is combined with the output component corresponding to the fundus reflected light corresponding to the fundus reflected light. It is necessary to separate the output component into the output component corresponding to the first Purkinje image forming reflected light corresponding to one Purkinje image PI.

This is because the actual light amount distribution is as shown by the solid line GS in FIG. 16, and is processed without being separated into the output component G6 corresponding to the fundus reflected light and the output component G7 corresponding to the first Purkinje image forming reflected light. If so.

The center of gravity position II (coordinates or address) including both of these is determined, and the center 34 of the pupil and the first Purkinje image P
This is because the center of I cannot be found.

In this case, in order to separate the output component G6 corresponding to the fundus reflected light and the output component G1 corresponding to the first Purkinje image forming reflected light as accurately as possible, it is necessary to set the slice level SL near the boundary between them. .

For this purpose, a plurality of zone levels ZN are provided and the output frequency of the photoelectric conversion element 53a is checked.

Here, it is assumed that there are eight zone levels ZN as shown in FIG. 17. Note that these eight zone levels ZN are indicated using symbols 211 to ZN.

In order to check the output frequency of the photoelectric conversion element 53a, eight output frequency registers R, .about.R, are prepared in correspondence with the eight zone levels ZN, .about.ZN. Note that the number of bits of the output frequency registers R, .about.R, is 8. Then, the outputs of the photoelectric elements 53a from addresses a to b are sequentially input into the output frequency registers R□ to R. For example, since the output of address a is "0",
The contents of all appearance frequency registers are "0". Now, i
When the output of the photoelectric conversion element 53a at the address is an output corresponding to "21", the content of the appearance frequency register R is "00000010J, and the content of the other appearance frequency registers is rOJ. Also, for example, i+1 When the output of the photoelectric element 53a at the address is larger than the output "zllll" of the photoelectric conversion element 53a at the i address by an amount corresponding to one bit, the content of the appearance frequency register R3 is rlooo.
It becomes ooloJ.

Therefore, we focused on the upper 3 bits of the output frequency registers R1 to R, and determined that the data in the upper 3 bits was at least [1
", the output frequency register R1-R contains "
+IJ is output. and.

The output of the photoelectric element 53a at each address (i = a to b) is input, and each time the contents of the upper three bits include "1", the output of each appearance frequency register R1 to R is incremented. Note that the content of the upper 3 bits is “1”.
", the increment count is not performed. In this way, each time the photoelectric element 53a at each address outputs, the appearance frequency registers R□ to R, are incremented. In the case of the output distribution shown schematically, Since the number of photoelectric elements 53a having an output level between zone level ZN2 and zone level ZN is the largest, it is expected that the increment count number of appearance frequency register R3 will be the largest.

Therefore, after incrementing the output distribution of the photoelectric cable 53a at all addresses, the appearance frequency register R1
It is determined whether the increment count number of ~R has reached the maximum. Then, the zone level ZN corresponding to the output frequency register R, to R, whose increment count is the maximum is determined as the slice level SL. By using this slice level SL, it is possible to separate the output component G6 corresponding to the fundus reflected light and the output component corresponding to the first Purkinje image forming reflected light.

Here, the width of the zone levels ZN, ~ZN, is determined according to the noise level based on the reflection from the fundus, and this noise level component can be removed through a low-pass filter, but the width of the zone levels ZN1 ~ ZN.

This can also be done by software processing of overlapping.

For example, as shown in FIG. 18, the sum of the increment counts of adjacent output frequency registers R□-R is calculated, and the output frequency register R□-R with the largest sum is determined. In the example shown in FIG. 18, the appearance frequency register R1 and the output frequency register R.

Since the sum is the maximum, it is determined that the increment count number of the appearance frequency register R4 is the maximum.

Note that among the output components G corresponding to fundus reflected light, the output component that appears most frequently is at an intermediate level.

Regarding the determination of slice level SL, zone level ZN,
, ZN, corresponding to the frequency register R,. Consider excluding R from the beginning.

Assume that it is possible to obtain the zone level ZN corresponding to the appearance frequency register R, in this way. here,
The contents of the appearance frequency register R, are.

It is predetermined in advance that when the value is equal to or more than roooooooolJ, the output component G corresponds to the first Purkinje image forming reflected light, and when it is equal to or less than "r00000110", the output component corresponding to the fundus reflected light G6.

In this way, the appearance frequency register R.

As shown in FIG. 16, based on the contents of
This means that it can be set near the boundary with the Purkinje image forming reflected light corresponding output component G7.

In this way, by determining the slice levels SL□, SL, and performing image separation processing by slicing the output components corresponding to the light intensity distribution characteristics shown in FIG. 16, the separated output shown in FIG. 19 is obtained. . In this No. 1911, solid line G,
indicates the separated output corresponding to the fundus reflected light.

A solid line G indicates the separated output corresponding to the first Purkinje image forming reflected light. Here, the separate output G8 corresponding to the fundus reflected light has a trapezoidal shape, which means that the output of the one-dimensional line sensor 53 is divided into the separated output G corresponding to the fundus reflected light and the separated output G corresponding to the first Purkinje image forming reflected light. This is because the above-mentioned correction process was performed before separating into , and. Therefore, the position of the center of gravity of the separated output G corresponding to the fundus reflected light is X□, and the distance lid from the center 34 of the patent pupil to the first Purkinje image, which is characterized by the position of the center of gravity of the separated output G corresponding to the reflected light forming the first Purkinje image, is
' is determined as +d=X, -X, -The calculation algorithm for determining the center of gravity position is PSD.
(position sensor diode) output realized by software calculation is used. That is, as shown in FIG. 20(a) and FIG. 20(b), using the weighting functions WA, W, the convolution of the image separation output corresponding to the output of the weighting functions WA, W, After taking the convolution integral, this is integrated. for example,
Image separation and output G shown in Fig. 20(C) and Fig. 20(d)
1 and the weight function W, , W, is the conpolution 3,
and obtain the multiplication output CA,C. And this multiplication output C. , C, to obtain the integral value S,,S,.

Then, the center of gravity [X is the distance from the origin 0 as Sf.

X=Sr* ((SA-S,)/(SA+S,)+1)
This method requires a bitwise multiplication to perform the convolution. Recently, it has become common for microcomputers to have a multiplication function, so the position of the center of gravity can be determined using this method.

However, if this center of gravity position iiix is determined using software realignment, there is a disadvantage that the calculation takes too much time.

Therefore, a processing means that can calculate the center of gravity Ifx while reducing the calculation time is adopted here.

First, the obtained separated outputs G,,G, are bit-inverted with respect to the position coordinates, and as shown in FIG. 19, the inverted branch outputs G,
-G, is generated.

According to this method, the separated outputs G, , G before inversion.

By calculating the phase difference between the separated outputs G,', and Gg' after inversion, the center of gravity position can be determined with approximately the same accuracy as the above. It can be determined by a calculation method similar to the correlation method calculation of the phase difference detection method used in a single-lens reflex camera having an autofocus optical system. Furthermore, this calculation method is as follows.

By interpolation calculation, the resolution of the sensor pixel can be calculated from number lO to number 10.
It has been known for some time that it can be obtained with an accuracy of 1/0.

By the way, unlike when photographing a subject that is completely unpredictable, this line-of-sight direction detection device [46]

The pattern of the obtained image is predictable, and when the reflected light from the fundus of the eye and the reflected light forming the first Purkinje image PI are imaged spot-wise on the one-dimensional line sensor 53, they are symmetrically separated. Outputs G,,G, are obtained. Therefore, for example, as shown in FIG. 21, when the separated output G,' has a simple pattern, the center 0 between the rising position coordinate and the falling position coordinate is expected to be approximately the center of gravity. Therefore, when detecting the phase difference.

If we perform calculations only around the center 0@.

The calculation time can be shortened.

Specifically, the output of the one-dimensional line sensor 53 is expressed as S(n)
shall be. Here, n indicates the address of the photoelectric element 53a of the one-dimensional line sensor. Then, focusing on address n and address n+1, a difference output E(n) of the separated outputs is generated. The difference E(n) is determined by the following formula.

E(n)=S(n+1)-5(n) In this way, a differential output B1 as shown in FIG. 21 is obtained.

Next, if the coordinates where E(n) is maximum and minimum are t□ and t2, respectively, then the center of gravity position is.

It can be expected to be approximately (t, +t,)/2.

Therefore, the inverted separation output when the position coordinates are inverted is G
,'' and generate the difference output R(n).

The differential output B,' corresponding to this difference output R(n) is as shown by the solid line. Here, as the total number of bits m,
The position of the center of gravity can be determined by performing a correlation method calculation to determine the phase difference of R(n) with respect to S(n) for a pair before and after m-(t, +ty).

Similarly, the phase difference between B and B6' can also be determined.

That is, if the phase difference between R (n) with respect to S (n) or B and B,' is t, then S (n)
The center of gravity position from the center coordinates 06' of the sensor can be determined by t/2.

By using such a calculation algorithm, a highly accurate line-of-sight direction detection device can be realized.

By the way, B1 and B', unless you adopt the method of calculating the phase difference, -R(n) corresponds to the address of the memory where S(n) is stored, so in reverse order from the address If the data is called, there is no need to create a memory area for generating R(n), and memory can be saved.

Also. The purpose of generating E(n) is to find the maximum and minimum addresses, not to obtain E(n), so the generation area is not necessary.

By the way, what about the gaze direction detection device in the previous example? The optical system of 146 includes a pentaprism 40 and a finder loupe 44.
Since the light transmitting system 46A and the light receiving system 46B were built into the camera body on the opposite side, the light transmitting system 46A and the light receiving system 46B
Reflected light based on the refractive surface of each optical element constituting the optical element is guided as a ghost to the light receiving system 46B, and a ghost is formed along with the first Purkinje image PI on the one-dimensional line sensor 53 of the light receiving system 46B, and the ghost and the first Purkinje image The problem remains that it is difficult to distinguish it from PI.

Therefore, next, an optical system of a camera line-of-sight direction detection device that prevents ghosts from being guided to the light receiving system 46B as much as possible will be described.

31 to 35 are explanatory diagrams of the optical system of the camera line-of-sight direction detection device in which the ghost is prevented from being guided to the light receiving system 46B as much as possible, and the components of the optical system shown in FIG. Identical components are given approximately the same reference numerals.

Here, the light transmission system 46A includes a light source 48 that generates infrared light.
．． It is equipped with a total reflection mirror 149 and a collimator lens 150. The surface A of the collimator lens 150 is an aspherical surface. The infrared light emitted from the light source 48 is reflected by a total reflection mirror 149 and guided to a collimator lens 150. On the exit side surface of this collimator lens 150,
A diaphragm 151 is provided. collimator lens 150
has a function of converting the infrared light emitted from the light source 48 into a parallel light beam.

A coaxial forming optical member 152 is provided on the side of the finder magnifying glass 44 facing the eye 45 to make the optical axis of the light transmitting system 46A and the optical axis of the light receiving system coaxial. This coaxial forming optical member 152 is used here.

It is composed of a rectangular parallelepiped made up of prisms 154 and 155 having a reflective surface 153. The coaxial forming optical member 152 has a transmission surface 156 facing the eye 45 and a reflection surface 153.
It has a transmitting surface 157 that faces the transmitting surface 156 and a transmitting surface 157' facing the collimator lens 150.

A mask 15g is provided on the transmission surface 156.

Here, in order to avoid ghosts due to reflection on each transmission surface of the coaxial forming optical member 152, the transmission surface 15
6.157 is tilted very slightly with respect to the optical axis,
The transmission surface 157' is tilted very slightly with respect to the optical axis. Each optical axis 2. , each transmission surface 1 for jlL
The inclination angle of 56.157.157 is +
t'', and since each transmission surface 156, 157, 157' has the same inclination angle, the state is the same as that in which a parallel plane plate is inserted, and there is almost no change in aberration due to inclination.

Here, the reflective surface 153 is semi-transparent for infrared light and transparent for visible light. Since the reflective surface 153 transmits visible light, the photographer can see the subject image formed on the one-piece plate 42. The parallel light beam passing through the aperture 151 is reflected by the reflecting surface 15
3 in the direction toward the eye 4, and is projected onto the photographer's eye 45 placed at the eye point. In this embodiment, the coaxial forming optical member 152 is used, but a mirror that semi-transmits infrared light and transmits visible light may also be used.

A corneal specular reflection light beam forming a first Purkinje image PI;
The reflected light flux from the fundus is again the coaxial forming optical member 15.
2, passes through the reflective surface 153, and is guided to the finder magnifying glass 44. The finder magnifying glass 44 has a device light receiving system 46B composed of lenses 44a and 44b as described above, which here includes a compensator prism 1.
59, a reduction lens 50, a total reflection mirror 161, a re-imaging lens 52, and a one-dimensional line sensor 53. The re-imaging lens 52 has a. As shown in an enlarged view in FIG.
It is located on the side facing the street.

By the way, also in this example, it is preferable that no distortion exists in the light receiving system 46B, and in relation to the object height, it is preferable that the light amount distribution on the one-dimensional line sensor 53 is approximately uniform. When the optical system is configured as described in .

As shown in #34@, within the required object height range,
The light intensity distribution on the one-dimensional line sensor 53 can be made substantially uniform, and the distortion can be made less than 1 μm as shown in FIG. 35.

(1) Design value of the light transmitting system 46A Radius of curvature of the exit surface of the light source 48...U of infinite size
Huang Ken's first saliva with Jin Shinju and Kinryoku 8sha Mirror 149...
7.7am total reflection mirror 149 and collimator lens 1
Distance from surface A of 50 - 7.3s Collimator lens 15
0 Radius of curvature of surface A...to. oo radius of curvature of surface B 1・-28.00 m refractive index ---1.48304 center thickness --- 4.00 light s between abomination mask 151 and surface B of collimator lens 150
s+・4. (X) One mask 151 Thickness...-0.04■■ Radius of curvature...Distance between optical axes between infinite mask tSt and transmission surface 156'...-0
．． 66m■11] y Two radius of curvature - Inclination with respect to the infinite optical axis -1.

Refractive index of optical member 152 for coaxial formation...-1.50 Jl
71 Optical axis distance layer from transmitting surface 157' to transmitting surface 156 ---lz- ■ path 1 and 156 Radius of curvature --- Infinite optical axis. Inclination...1 Optical axis distance III from the transmission surface 156 to the cornea 32-...
・1355 Peng Cornea 32 radius of curvature...-7.980 ■■
Note that the surface A of the collimator lens 150 is an aspherical surface,
In the imaging formula for an aspheric lens described below.

k=-3.165, a,=-2.95XIO-. a.
= 0, the sag amount X was determined and designed.

X=(a,h+a.h)+c-h”/(1+J
l −(k+ 1)♂·h2) Note that C is the reciprocal of the radius of curvature of the surface A of the collimator lens 150, h is the height from the optical axis, and k is the aspheric coefficient.

(2) Radius of curvature of design angle a32 of light receiving system 46B - -7.980+am cornea 3
Distance between optical axes from z to transmission surface 156...l3■■1
111 Inclination with respect to the optical axis...--1" Radius of curvature...-Refractive index of the optical member 152 for forming an infinite common ring...-1.5087
1 What optical axis distance between the transmitting surface 156 and the transmitting surface 157...l 抛■Transmitting surface 157 Inclination with respect to the optical axis...--t'' Radius of curvature--From the infinite transmitting surface 157 to the surface A of the lens 44a Distance between optical axes: 0.60 m Lens 44a Radius of curvature of surface A: 115.895 Layer center thickness:
1.2 ■ Otori refractive index -1,69747 Radius of curvature of surface B -29,210 - ■ Lens 44b Radius of curvature of surface B - 2L210 ■ Thickness at the center of Peng ... -4,92 Otori - Refractive index -1,61187 Radius of curvature of surface C...-47,880-Optical axis distance between surface C and surface A of pentaprism 40...1.00+w Radius of curvature of surface A of pentaprism 40... Infinity refractive index --- 1.50871 Radius of curvature of surface B --- Inclination of surface B with respect to infinite optical axis aJ ----24° Distance between optical axes from surface A to surface B --- 28. 80■ Distance between optical axes between layer surface B and surface A of compensator prism 159...-0
-14 ■ Compensator prism 159 Radius of curvature of surface A ---Inclination of surface A with respect to the infinite optical axis - -1-24° Radius of curvature of surface B --- Optical axis of infinite surface A and surface B Distance: 3 ■■ Refractive index: 1
．． 50871 Distance from surface A to mask 159' -0 ■■ Mask 1
59' Thickness...-0,04■■ Radius of curvature--Distance between optical axes from the infinite mask 159' to surface A of the reduction lens group--0,10m Radius of curvature of surface A of the reduction lens 50- -・11.71 Raw meal thickness-・・2.
50 ■■ Radius of curvature of surface B ---60,140 Garden refractive index ---
1.48304 Distance between optical axes from surface B to total reflection mirror 161 - 3
．． 00mm Radius of curvature of total reflection mirror 161... Distance between optical axes from infinite total reflection mirror 161 to re-imaging lens 52 = 1.61)m Radius of curvature of surface A of re-reflection lens 52...-1, 520 layers ■ Refractive index - 1,48304 Center wall thickness of the enteret - 1.520 m Radius of curvature of layer surface B - Distance from infinite surface B to mask 54 - 0.00 Iris mask 54 Radius of curvature -... Infinite thickness = 0.04 mB Note that the surface A of the reduction lens 50 is an aspherical surface, and in the above formula, K = -1,25,a, = -8X 10-
, α, = −10−.

FIGS. 36 to 38 are diagrams for explaining a second embodiment of the line-of-sight direction detection optical system of a camera according to the present invention,
In this embodiment, the light transmitting system 46A is provided on the opposite side of the finder loupe 44 with the pentaprism 40 interposed therebetween, and the light receiving system 46A is provided on the opposite side of the finder magnifying glass 44.
6B is provided on the transmitting surface 157' side of the common ring forming optical member 152, and the infrared light emitted from the light source 48 is transmitted to the convencator prism 159.6B. Through the pentaprism 40,
The infrared light is guided to a finder magnifying glass 44, which converts the infrared light into a parallel beam of light and projects it onto the eye 45.
The light beam forming the Purkinje image PI and the reflected light from the fundus are reflected by the reflective surface 153 of the coaxial optical member 152 and guided to the light receiving system 46B,
The other optical components are almost the same as those of the first embodiment, and the optical characteristics are also almost the same as those of the first embodiment, as shown in FIGS. 6 and 7, so the design will be described below. Just state the value.

(1) Design value of the light transmission system 46A Radius of curvature of the output surface of the light source 48...Distance between the optical axes between the output surface of the infinite light source 48 and the total reflection mirror 149...17■
Radius of curvature of total reflection mirror 149 - Optical axis distance between infinite total reflection mirror 149 and mask 159' - 31 ■ Mask 159' Thickness: 0.04 Radius of curvature: Infinity mask 15Q Distance between surface B of compensator prism 159...7, - Compensator prism 159 Radius of curvature of surface B...infinity Distance between surface A and surface B...3 ■■ Radius of curvature of surface A...infinity Inclination of surface A with respect to large optical axis deviation -24° Distance between optical axes between surface A and surface B of pentaprism 40 ...
0,14 ■■ Pentaprism 40 Radius of curvature of surface B...Inclination of surface B with respect to infinite optical axis i -24° Refractive index -1,50871 Radius of curvature of surface A... From infinite surface A Distance between optical axes to surface B -28,80mm Distance between optical axes between surface A and surface C of lens 44b...1.00-■
Lens 44b Radius of curvature of surface C...-47,880 Radius of curvature of surface B...-29,210■■Center thickness...4.92K Refractive index...1.61187 Lens 44a Radius of curvature of surface B...-29,210 Radius of curvature of surface A=--115,895 mm Center thickness...-1,2 Refractive index...1.69747 Surface A and transmission surface 57 Which distance between optical axes...0.60 ugly m stealth surface 157 Radius of curvature...infinity optical axis, inclination...2° Refractive index of optical member 152 for forming coaxial axis...-1, 5087
1 Distance between optical axes from transmission surface 1-57 to transmission surface 156 -
10m ■Transmission surface 156 Radius of curvature...infinite optical axis 2. Inclination...2° Distance between optical axes from the transmission surface 156 to the cornea 32...13
Peng - Radius of curvature of cornea 32...7.980+++n
(2) Design value of light receiving system 46B Radius of curvature of cornea 32...-7,980 ■■ Distance between optical axes from cornea 32 to transmitting surface 156...13 mm Transmitting surface 156 Radius of curvature...-Infinite light Inclination with respect to axis...-2° Distance between optical axes from transmission surface 156 to transmission surface 157'--
-12mm refractive index of optical member 152 for coaxial formation...1
1508711111 Company' Inclination with respect to the optical axis - -2° radius of curvature - Distance between the optical axes from the infinite transmission surface 157' to the mask 151...
・0.66 Distance between dragon mask 151 and reduction lens 50...0.00■■Mask 151 Radius of curvature...infinite thickness...-0.04 Curvature of surface A of reduction lens 50 between parts Radius: 28,00 Thickness: 4.00 - Radius of curvature of surface B: 10,00 Refractive index: 1
．． 48304 Distance between optical axes from surface B to total reflection mirror 161...-7
, curvature radius of the 30 mm total reflection mirror 161... distance between the optical axes of the infinite total reflection mirror 161 and the surface A of the re-imaging lens 32 -5, curvature radius of the Hu re-imaging lens 52 surface A -2 ,00■■ Refractive index...-1,48304■- Center thickness...2.00mm- Radius of curvature of surface B...Distance from infinite surface n to mask 54...0.00m Mask 54 Radius of curvature... Infinite thickness - -0.04 m Note that the surface B of the reduction lens 50 is an aspherical surface, and the above formula is satisfied, K = -3,165, a, = 2. 95
It was designed as X 10−−α6=O.

According to this line-of-sight direction detection device, it is possible to avoid ghosts in the light receiving section as much as possible.

As explained above, the line-of-sight direction detection device for a camera according to the present invention has a light transmitting system that guides a parallel light beam to the photographer's eye, and a light receiving section, and is capable of detecting the specular reflection of the cornea of that eye. a light-receiving system that receives the reflected light that forms the first Purkinje image and the reflected light from the fundus of the eye; and a processing circuit that detects the line-of-sight direction of the photographer's eye based on the light-receiving output of the light-receiving section. Since it is provided on the camera body, it has the effect of detecting the line of sight direction of the photographer's eyes when looking into the camera.

Further, in a camera in which a finder is provided with a plurality of focusing zones, the autofocus optical system corresponding to the focusing zone can be automatically selected and driven.

[Brief explanation of the drawing]

1 to 5 are for explaining an example in which the line-of-sight direction detecting device according to the present invention is applied to a single-lens reflex camera. FIG. 1 is an explanatory diagram showing how a line-of-sight direction detecting device according to the present invention is arranged on a camera, and FIGS. 2 and 3 are detailed views of the line-of-sight direction detecting device. FIG. 4 is an enlarged view of the re-imaging lens shown in FIGS. 2 and 3. FIG. 5 is a schematic diagram of the line-of-sight direction detection device. FIG. 6 is a graph of spherical aberration when the reduction lens shown in FIGS. 2 and 3 is not made into an aspherical surface. Figure 7 is a graph of distortion when there is the spherical aberration shown in Figure 6, Figure 8 is a graph of spherical aberration when the reduction lens shown in Figures 2 and 3 is made into an aspherical surface, Figure 9 is a graph of distortion when there is no spherical aberration shown in FIG. FIGS. 11 and 11 are schematic diagrams showing the relationship among the line-of-sight direction detection device, reimaging lens, finder loupe, photographer's eye, and one-dimensional line sensor of a camera according to the present invention. Figures 12 and 13 are schematic diagrams for explaining problems when using a one-dimensional line sensor as a light-receiving element in the line-of-sight direction detection optical system. Figure 14 is a diagram showing the amount of light at the periphery of the re-imaging lens. FIG. 15 is an explanatory diagram of a correction processing means for correcting attenuation, and FIG. 15 is a block diagram of a processing circuit having the correction processing means. Fig. 16 is a schematic diagram showing the relationship between the actually obtained light intensity distribution and the one-dimensional line sensor, Figs. 17 and 18 are explanatory diagrams of the image separation processing means, and Fig. 19
21 to 21 are explanatory graphs for determining the center of gravity position of the image separation output distribution, and FIGS. 22 to 24 are explanatory diagrams for explaining the detection principle of the line-of-sight direction detection device according to the present invention. , Fig. 22 is an explanatory diagram showing the state in which a light spot is formed when a parallel light beam is irradiated onto a convex mirror, and Fig. 23 is a state in which a first Purkinje image is formed when a parallel light beam is irradiated on the cornea of the eye. An explanatory diagram showing. Figure 24 is an enlarged view of the eye to explain the relationship between the first Purkinje image and the center of the pupil. Figures 25 and 26 are for calculating the line of sight direction of the eye from the first Purkinje image and the center of the pupil. Fig. 27 is a perspective view schematically showing the layout of the improved autofocus optical system of a single-lens reflex camera, and Fig. 28 is an explanatory diagram for determining the focus of the single-lens reflex camera by focusing the photographing lens on the center of the finder. An explanatory diagram for explaining the relationship between the exit pupil and the aperture area as seen from the focusing zone of the autofocus optical system, which is optically approximately conjugate with the river zone. Figure 29 is a plan view of the finder of the single-lens reflex camera. . FIG. 30 is an explanatory diagram for explaining the relationship between the exit pupil and the aperture area when the exit pupil shown in FIG. 27 is subjected to vignetting. 31 to 35 are diagrams for explaining still other examples of the optical system of the line-of-sight direction detection device according to the present invention, and FIG. 31 is a configuration diagram of the optical system of the line-of-sight direction detection device; 32 is an enlarged view of the main parts of the optical system of the line-of-sight direction detection device shown in FIG. 31, and FIG. 33 is an enlarged view of the re-imaging lens shown in FIG. 31. 34 and 35 are explanatory diagrams of the optical characteristics of the optical system of the line-of-sight direction detection device shown in FIG. 31. 36 to 38 are diagrams for explaining other examples of the optical system shown in FIG. 31, and FIG. 36 is an optical diagram showing the main part configuration of the optical system of the line-of-sight direction detection device; 37 and 38 are explanatory diagrams of the optical characteristics of the optical system shown in FIG. 36, and FIG. 39 is a diagram showing a schematic configuration of an autofocus optical system of a conventional single-lens reflex camera. FIG. 40 is a perspective view schematically showing the arrangement of the autofocus optical system shown in FIG. 39. FIG. 41 is an explanatory diagram for explaining focusing by the autofocus optical system. FIG. 42 is an explanatory diagram of the detection output of the CCD of the autofocus optical system, and FIG. 43 is an explanatory diagram for explaining the arrangement of conventional focusing zones in the finder. FIG. 44 is an explanatory diagram for explaining the photographing procedure when using the conventional single-lens reflex camera to obtain a photograph in which a desired subject is shifted from the center to the left or right. It is. 9... Autofocus optical system, 16... Finder 17... Central focusing zone 18.19... Autofocus optical system for peripheral focusing 26.27... Peripheral focusing zone 28.29 -- Focusing zone 32 -- Cornea, 4G -- Pentaprism 34 --
- Center of pupil 44...Finder magnifying glass, 45--Photographer's eye 46...Line-of-sight direction detection device 46A--Light sending system, 46B--Light receiving system, 48...
Infrared light source 50...Reducing lens, 52...Re-imaging lens 53--One-dimensional line sensor, 55... Aperture 58
... Microcomputer 152 - Optical member for coaxial formation 156.15... Transmission surface, countersunk, flA - Optical axis 53a - Photoelectric element, 0... Rotation angle, 5%...
・Center of rotation PI--・First Purkinje image, G, -G,・
・・Light amount distribution H, ・・・Correction coefficient,
,...-separation output corresponding to fundus reflected light-G,...separation output system corresponding to first Purkinje image forming reflected light 1 7 1s, 1 yen! If I
l〒! Shitaku 4'' 54, 口152 Taku 5 Figure 2 5W#I$ Book List No. 6 - 100to. Taku 8 figure enemy ＼-0,034 -0,100,10 7th: @/@-t″A stoshi reference I 2 t, 6mm 1 mm view 10= l rabbit 12 mouth and tail 14 figure X ,,, Q. 53a l, 53 1'': 1 j l j +, -x, = x, 1, 1 1 1 '1 tea 15 = -4: lRoMli + + 1 o, -% no, and G5 1st 18; - Enter 6s. (□ Q'jx~・0;-Qft"-Naki", 1 ..1 \1 /■ 21st i 1 /7! IS. +1jjljl1 1, : lli8-BE(E<n)):: i Ou-AoE Fig. 22 Q7 Do\\730 Fig. 25 Fig. 26! cI + +: ii Dona. Fig. 29 Fig. 28 Fig. 30 Taku 31 ward: 58 1 iL-Ji 1 3rd
4 part No. 37 figure disc 1 (#J) 10 has No. 39 copy 40 parts

Claims (18)

[Claims] (1) A light transmitting system that guides a parallel light beam to the photographer's eye, and a light receiving unit that includes reflected light that forms a first Purkinje image based on corneal specular reflection of the eye and reflected light from the fundus of the eye. and a processing circuit for detecting the line-of-sight direction of the photographer's eye based on the light-receiving output of the light-receiving section, the camera body being provided with: Device. (2) The camera according to claim 1, wherein at least one of the light transmitting system and the light receiving system is incorporated into the camera body on a side opposite to the finder magnifying glass with a pentaprism as a boundary. Gaze direction detection device. (3) The line-of-sight direction detection device for a camera according to claim 1, wherein the parallel light beam is infrared light. (4) The light transmitting system includes an infrared light source that generates infrared light that is emitted as a parallel beam of light toward the photographer's eye via a finder loupe, and the light receiving system includes the corneal specular reflection. 2. The line-of-sight direction detection device for a camera according to claim 1, further comprising a reduction lens that reduces the reflected light forming the first Purkinje image and the reflected light from the fundus of the eye to form an image. (5) At least one of the reduction lenses has an aspherical surface, and the light receiving system is provided with a re-imaging lens that re-images the reflected light forming the first Purkinje image, and the re-imaging lens 5. The line-of-sight direction detection device for a camera according to claim 4, wherein an aperture is provided at a center of curvature of the reduction lens, and a focal point of the reduction lens is located at a center of curvature of the re-imaging lens. (6) The camera body is provided with a plurality of focusing zones within the field of view of the finder, and an autofocus optical system corresponding to the focusing zone is provided at a position substantially optically conjugate with the focusing zone. 2. The camera according to claim 1, wherein the camera is provided with focusing zones of the viewfinder, and the processing circuit automatically senses that any one of the focusing zones of the finder is selected. Gaze direction detection device. (7) The processing circuit is connected to a drive circuit that drives an autofocus optical system corresponding to a selected focusing zone among the focusing zones of the finder. A camera line-of-sight direction detection device described in . (8) The light-receiving system includes a re-imaging lens that re-images the reflected light forming the first Purkinje image based on corneal specular reflection on the light-receiving section, and the light-receiving section is connected to the plurality of autofocus optical systems. A mask having an opening is provided between the re-imaging lens and the one-dimensional line sensor, and a mask having an opening is provided between the re-imaging lens and the one-dimensional line sensor. 2. The line-of-sight direction detection device for a camera according to claim 1, wherein the lens is a cylindrical lens that forms an elongated image in a direction perpendicular to the arrangement direction of the photoelectric elements of the one-dimensional line sensor. (9) The light-receiving system includes a re-imaging lens that re-images the reflected light forming the first Purkinje image on the light-receiving section based on corneal specular reflection; It is composed of a one-dimensional line sensor having photoelectric conversion elements arranged corresponding to the focusing zone, and a mask having an opening is provided between the re-imaging lens and the one-dimensional line sensor, and the re-imaging 2. The line-of-sight direction detection device for a camera according to claim 1, wherein the lens is a toric lens that forms an elongated image in a direction perpendicular to the arrangement direction of the photoelectric elements of the one-dimensional line sensor. (10) The light receiving section is composed of a one-dimensional line sensor, and the processing circuit processes the output from the one-dimensional line sensor at one slice level to obtain coordinates corresponding to the pupil periphery, and , obtain Purkinje image corresponding coordinates corresponding to the first Purkinje image by processing at another slice level, calculate the center coordinates of the first Purkinje image and the center coordinates of the pupil, and detect the line of sight direction of the eye. The line-of-sight direction detection device for a camera according to claim 1. (11) The light receiving section is composed of a one-dimensional line sensor, and the processing circuit converts the output from the one-dimensional line sensor into a fundus reflected light corresponding output component corresponding to the reflected light from the fundus and forms a first Purkinje image. the center of gravity of the separated output component corresponding to the fundus reflected light and the center of gravity of the output component corresponding to the first Purkinje image forming reflected light; 2. The camera's line-of-sight direction detection device according to claim 1, wherein the camera's line-of-sight direction is detected by determining the position and position of the eye, respectively. (12) The light receiving system includes a re-imaging lens that re-images the reflected light forming the first Purkinje image based on corneal specular reflection on the one-dimensional line sensor, and the processing circuit includes a re-imaging lens that re-images the reflected light forming the first Purkinje image based on corneal specular reflection, 12. The line-of-sight direction detection device for a camera according to claim 11, further comprising a correction means for correcting a decrease in the amount of light incident on the peripheral portion based on the light amount distribution characteristics. (13) The separated output component corresponding to the fundus reflected light and the first
13. The line-of-sight direction detection device for a camera according to claim 12, wherein the position of the first Purkinje image and the position of the pupil are determined by bit-inverting the output component corresponding to the Purkinje image forming reflected light. (14) A light transmitting system that emits detection light as a parallel light beam toward the eye looking through the finder magnifying glass, and a light receiving system that reimages the detection light that forms a virtual image on a light receiving unit based on specular reflection of the cornea of the eye. , characterized in that a coaxial forming optical member for making the optical axis of the light transmitting system and the optical axis of the light receiving system coaxial is provided on the side of the finder magnifying glass facing the eye. A camera line-of-sight direction detection device. (15) The light receiving system includes a reduction lens and a reimaging lens between the coaxial forming optical member and the light receiving section,
The line-of-sight direction detection device for a camera according to claim 14, wherein at least one surface of the reduction lens is an aspherical surface. (16) The camera according to claim 14, wherein the coaxial forming optical member is a mirror that transmits light in the visible region and has characteristics of reflecting and transmitting light in the infrared region. line of sight direction detection device. (17) The line-of-sight direction detection device for a camera according to claim 16, characterized in that a prism having a reflective surface is used in place of the mirror. (18) The prism includes a transmitting surface facing the eye and a transmitting surface facing the Van Ider magnifying glass and facing the reflecting surface narrowly, the transmitting surface facing the eye being at least slightly opposite to the co-axis. 18. The viewing direction detection device for a camera according to claim 17, wherein the device is tilted at an angle of .