JP3222164B2 - Imaging device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical system which forms an image on an image pickup surface of an image pickup device which samples light intensity on the image pickup surface and picks up an image.

[0002]

2. Description of the Related Art In recent years, an image pickup optical system using a solid-state image pickup device such as a CCD has been used as a device for picking up an image to be observed on a television monitor. An imaging element used in these imaging devices samples the light intensity distribution on the imaging surface at a certain frequency, converts the light intensity into an electric signal, and reproduces it on a television monitor.

However, when an image is reproduced by sampling as described above, only the frequency up to the Nyquist limit can be reproduced by the sampling theorem. Further, near the Nyquist limit, a pseudo-resolution called moire occurs.

For this purpose, various measures have been taken so that an optical system has a low-pass effect so that an image having a frequency near the Nyquist limit is not resolved on the imaging surface. For example, there are well known methods of creating a double image using a quartz filter and blurring the image using a phase filter.

A method of forming a double image by allowing a part of a light beam to pass through a wedge prism as disclosed in Japanese Utility Model Laid-Open No. 63-24523, Japanese Patent Application Laid-Open No. 47-38001, etc. And a method of forming an annular image with a conical prism as described in Japanese Patent Application Laid-Open No. 63-6520, and blurring an image by generating high-order spherical aberration as disclosed in JP-A-63-6520. There are methods.

[0006]

SUMMARY OF THE INVENTION In each of the above methods,
The method using a crystal filter which is considered to be most frequently used at present has a good low-pass effect because the point image is separated into two points. However, since the cutoff in the frequency space is one-dimensional, usually a plurality of quartz filters are used, which requires a lot of space and is very expensive. Particularly, when a sampled image such as an image guide is formed on an image sensor, a large number of crystal filters are required, so that the cost becomes higher.

Although phase filters are inexpensive, they generally have lower performance than quartz filters.

The method using a wedge prism has the same disadvantage as using a quartz filter because the point image is one-dimensionally separated.

The method utilizing higher order spherical aberration is relatively inexpensive and can provide a two-dimensional low-pass effect. However, rays near the optical axis are not significantly affected by higher-order aberrations. Therefore, high-frequency components cannot be completely removed, and it is difficult to obtain a good low-pass effect as compared with a quartz filter.

The conical prism has a two-dimensional low-pass effect, does not require a space like a wedge prism or a quartz filter, and furthermore, the light near the optical axis is separated by the prism. High frequency components do not remain unlike the method using aberration. However, the center diameter of the point image intensity distribution increases without substantially changing the minimum diameter of the circle of confusion due to defocusing, and as a result, the MTF increases, and a good low-pass effect cannot be obtained. Further, good point image separation cannot be obtained due to aberrations of the optical system.

An object of the present invention is to provide an image pickup apparatus provided with an optical system having a refracting surface which is inexpensive and has an excellent low-pass effect.

[0012]

Imaging apparatus of the present invention SUMMARY OF THE INVENTION is by sampling the light intensity on the image plane has an optical system for forming an image on the imaging surface of the imaging device for imaging the optical axis direction
Is z and r is the direction perpendicular to the optical axis, 2 in the equation (1)
At least one of the following terms is not 0 and
It is special to include refraction surfaces with different derivatives at the center and at the periphery.
Sign. z = C ₀ + C ₁ r + C ₂ r ² + C ₃ r ³ +... + C _n r ⁿ (C ₁ ≠ 0) (1) where r ² = x ² + y ² and r ≧ 0.

The apparatus according to the present invention has a structure as shown in FIG. 1, for example, 1 is an eyepiece which is a part of an endoscope including an eyepiece 1a, 2 is an adapter including the optical system, 3
Is a television camera, 4 is an imaging surface, and 5 is a cover glass. In the embodiment shown in FIG. 1, the above-mentioned refraction surface is formed on this cover glass. Reference numerals 6 and 7 are a quartz filter and an infrared cut filter, respectively.

The refracting surface provided in the optical system included in the apparatus of the present invention, which includes a point (or line) at which the differential coefficient is discontinuous, is disclosed in Japanese Patent Application Laid-Open No. 47-38001. Unlike a wedge prism or a conical prism, for example, as shown in FIG. 3, the refraction surface has a shape with different coefficients at the center and the periphery of the surface.

As shown in FIG. 15, in the prism disclosed in the above-mentioned publication, as shown in FIG. MTF rises to increase. In general, since the focus is adjusted to a position where the resolution is the best during the focus adjustment, a good cutoff cannot be obtained and moire occurs.

On the other hand, in the apparatus of the present invention, the refracting surface as described above is arranged so that a light beam passing through a portion near the optical axis and a light beam passing through a portion far from the optical axis as shown in FIG. 4, for example. Defocusing was caused by changing the imaging position, thereby preventing the MTF from increasing.

The surface shape as described above is generally represented by a combination of a plurality of polynomial aspheric surfaces. In some cases, the expression may be expressed by an aspherical expression including an odd order, and is expressed by the above expression (1).

In the above equation (1), when C ≠ 0, the condition is that the surfaces are not connected smoothly on the optical axis, that is, the surface contains a discontinuous point or line with a differential coefficient. Also C
_{If n} (n ≧ 2) is all 0, equation (1) is a straight line, but if one or more terms are not 0, it generally becomes a curve and the degree of freedom in designing to obtain an optimal separated image increases. preferable. Thus, the MTF of an image formed on the image pickup surface is near the Nyquist limit of the image pickup device, or provided with a sampling image transmission system such as an image guide. In the case where the image is formed on the imaging device of the sampling image transmission system, a cutoff may be made at any one of the vicinity of the frequency of the projected image on the imaging device.

The light beam split by the refracting surface forms a multiple image (or an annular image) on the image pickup surface, so that a low-pass effect works and moiré can be prevented. At this time, the condition for the annular image to cut off at the Nyquist limit is as follows. a ≒ 0.77p Here, a is the radius of the annular zone, and p is the sampling pitch of the image sensor.

Therefore, the inclination angle θ of the refracting surface near the optical axis
Can be represented by the following equation. | Θ | = a / {f ₂ · (n-1)} = {2 · f _n · f ₂ · (n-1)} ^-1 where n is the refractive index of the medium and θ is (A) in FIG. , (B) is the inclination angle near the optical axis of the refraction surface including the point (or line) where the derivative is discontinuous, which is usually a small angle, sin θ ≒ θ, f _n
Is the Nyquist limit (f _n = １ ／ p) of the image sensor or the sampling image transmission system, and f ₂ is the focal length of the optical system between the refraction surface including a point (or line) where the derivative is discontinuous and the image sensor. It is.

At this time, θ merely represents the relationship between the light near the optical axis and the Nyquist limit, and does not represent the cutoff frequency of the entire optical system. θ desirably satisfies the following equation (2). {0.2 ・ f _n・ f ₂・ (n-1)} ⁻¹ ≧ │θ│ ≧ {10 ・ f _n・ f ₂・ (n−1)} ⁻¹ (2) Since θ is usually small, In the case of an optical system having a small pupil diameter, the height of the protuberance on the surface or the depth of the depression becomes extremely small, which makes it difficult to manufacture, or the accuracy cannot be ensured due to variations. In such a case, if the refraction surface including a point (or line) at which the above-mentioned differential coefficient is discontinuous is set as a joint surface with a flat surface or a curved surface, (n-1) of the equation (2) becomes (n).
−n _b ), and the value of θ increases, so that the height of the protuberance of the surface or the depth of the depression can be increased, which is desirable. Note that n _b is the refractive index of the adhesive.

Also, between θ and C ₁ in equation (1),
Almost the following relationship holds. C ₁ = tan θ ≒ θ Further, if a refraction surface including a point (or line) where the above-mentioned differential coefficient is discontinuous is arranged in the vicinity of the pupil, a favorable low-pass effect works over the entire screen, which is desirable. At this time, the refracting surface including a point (or line) at which the differential coefficient is discontinuous or the prism (lens) including the refracting surface may be slightly eccentric with respect to the optical axis. That is, even if it is slightly eccentric near the pupil, it does not significantly affect the separation of point images,
This is because it also prevents ghosts.

A point (or line) at which the derivative is discontinuous
If the refraction surface including is rotationally symmetric, the MTF can be operated two-dimensionally, and the number of low-pass filters can be reduced.

When an image sensor having a different Nyquist limit between the horizontal direction and the vertical direction is used, an anamorphic low-pass filter as shown in FIG. It is effective to provide a cutoff near the Nyquist limit in the direction. In FIG. 6, when the optical axis direction is the z-axis, (A) is a yz section (vertical section) and (B) is an xz section (horizontal section). This anamorphic surface is represented by the following equation. _{z = A 0 + A 1 x} + B 1 y + A 2 x 2 + B 2 y 2 + A 3 x 3 + B 3 y 3 + ··· A n x n + B n y n (A 1 ≠ 0 or B ₁ ≠ 0) above refraction The surface may be rotated around the optical axis as needed.

In the case of a system in which a television camera is attached to a fiberscope and an image on a television monitor is observed,
Generally, since the image guide differs depending on the fiberscope, the frequency of the image guide formed on the image sensor of the television camera differs. Therefore, it is difficult to remove moiré from all the fiberscopes used.

In the above system, it is desirable to use an optimum low-pass filter according to the combination of the fiberscope and the television camera. For this purpose, for example, the adapter for connecting the fiberscope and the television camera can be exchanged, and the low-pass filter of the present invention that matches the frequency of the image guide can be arranged in the adapter. Further, a low-pass filter suitable for the image fiber may be arranged in the fiber scope.

When a rigid scope having a relay lens and a fiberscope are used in the same system, the use of a low-pass filter for the fiberscope in the rigid scope will lower the image quality more than necessary. Conversely, when a low-pass filter for a rigid endoscope is used in combination with a fiberscope, moire occurs.

Therefore, if the low-pass filter of the present invention can be replaced or the adapter including the low-pass filter of the present invention can be replaced, the rigid endoscope and the fiberscope can be properly used in the same system with optimum image quality. It becomes possible. Furthermore, if the moiré removal filter of the present invention disposed in the adapter or the television camera is replaceably disposed, an inexpensive and optimal system can be obtained.

The low-pass filter of the present invention can be used simultaneously with a plurality of filters, or can be used in combination with another low-pass filter such as a crystal filter. This is because the required low-pass effect differs depending on the system used.In particular, when an image sampled by an image guide or the like is formed on an image sensor, the sampling frequency is different. This is necessary. For example, in a CCD or the like with a color mosaic filter, the frequency at which moiré is emitted has directionality. Therefore, near the sampling frequency of the horizontal color where strong moiré appears, the MTF is dropped by the crystal filter, and the sampling frequency of the six-way dense image guide is in phase at intervals of 60 °, so that the MTF is reduced by the low-pass filter according to the present invention. Otose if an effective, expressed in almost the following equation the relationship between fiber distance φ and C ₁ of the image guide. 0.5 × φ × sin (π / 3) × β = C ₁ × f ₂ × (n−1) × 2 where β is a projection magnification on the image sensor. Also, as mentioned above, when the low-pass filter (refractive surface) is the joint surface,
In the above equation, (n-1) becomes (n- _nb ).

The low-pass effect, the value of C ₁ to receive the influence of the aberration of the optical system, it is not necessary to meet strict the formula, practically, it is sufficient to satisfy the following equation. 5 × φ × sin (π / 3) × β = C ₁ × f ₂ × (n-1) × 2 ≧ 0.1 × φ × sin (π / 3) × β Also, the low-pass filter of the present invention is integrated with the lens. It is desirable to further reduce the price.

[0031]

Next, examples of the present invention will be described.

FIG. 1 shows a configuration of an embodiment of the present invention. For example, an adapter 2 which is detachably mounted between an endoscope (eyepiece 1) and a television camera 3 is used for the endoscope. The mirror television adapter uses a rotationally symmetric low-pass filter having the cross-sectional shape shown in FIG. This FIG.
The low-pass filter shown in (1) is preferable because the surface other than on the optical axis is smooth, so that light scattering is small and flare hardly occurs.

A quartz filter having a trap line 10 as shown in FIG. 8 or 9 is arranged in the television camera 3. The low-pass filter of the optical element in the adapter is a first lens (cover glass), and the refraction surface of the low-pass filter has an aspherical shape including odd-numbered orders. As shown in FIG. 10, the trap line in the frequency space on the imaging plane by the low-pass filter matches the sampling frequency of the end face of the image guide projected on the imaging plane. In the figure, 11 is the sampling point of the image guide, and 12 is the top line.

Such a configuration is desirable because it can be combined with an adapter having an optimum cutoff frequency even when combined with an endoscope having a different sampling frequency. If the low-pass filter (cover glass) is replaceable, it is more preferable because one adapter can be used for a plurality of endoscopes.

The adapter of this embodiment has a zoom function, and the sampling frequency of the image guide on the imaging surface changes by zooming. In this case, if the low-pass filter is arranged before the variable-power optical system, the sampling frequency of the image guide on the imaging surface always coincides with the top line by the low-pass filter, so that moiré is less likely to occur at any magnification. System can do.

FIGS. 11 and 12 show MTFs at the wide end and the telephoto end, respectively, according to the first embodiment. However, the crystal filter is not considered.

The low-pass filters used in this embodiment are the same as those shown in FIG.
The following are used. Further, a low-pass filter as shown in FIG. 14 is also conceivable.

FIG. 2 shows the configuration of the second embodiment of the present invention.
This embodiment is an example in which the aforementioned refracting surface is provided after the stop in the objective lens of the television camera, and any type of low-pass filter shown in FIGS. 3, 6, and 7 may be used.

Next, data of the low pass filter used in the first embodiment and this embodiment will be shown. Example 1 f = 22.636 to 42.793, F / 4.9 to 9.3, object distance = -1000 r ₁ = ∞ (aperture) d ₁ = 0.484 r ₂ = ∞ d ₂ = 0.8 n ₂ = 1.51633 ν ₂ = 64.15 r ₃ = Ｄ d ₃ = 2.8 r ₄ = ∞ d ₄ = 1 n ₄ = 1.51633 ν ₄ = 64.15 r ₅ = ∞ d ₅ = 3.1 r ₆ = 13.71 d ₆ = 1.37 n ₆ = 1.72 ν ₆ = 50.25 r ₇ = -13.71 d ₇ = 1 n ₇ = 1.78472 v ₇ = 25.71 r ₈ = ∞ d ₈ = 5.134 to 7.703 r ₉ = -6.812 d ₉ = 1.5 n ₉ = 1.84666 v ₉ = 23.78 r ₁₀ = -3.705 d ₁₀ = 0.8 n _{10 = 1.62374 ν 10 = 47.1 r 11} = 8.719 d 11 = 7.527 ~1.251 r 12 = 18.929 d 12 = 2.76 n 12 = 1.62041 ν 12 = 60.06 r 13 = -13.442 d 13 = 0.2 r 14 = 9.197 d 14 = 4.71 n ₁₄ = 1.51633 ν ₁₄ = 64.15 r ₁₅ = -9.197 d ₁₅ = 0.8 n ₁₅ = 1.85026 ν ₁₅ = 32.28 r ₁₆ = 23.081 d ₁₆ = 2.171 to 5.879 r ₁₇ = ∞ d ₁₇ = 1 n ₁₇ = 1.51633 ν ₁₇ = 64.15 r ₁₈ = ∞ shape of refraction surface (r ₅ ) (A) [corresponding to FIG. Low-pass filter data] 0 <r ≦ 1, C ₀ = 0, C ₁ = 8.2 × 10 ⁻⁴ , 1 <r ≦ 2 C ₀ = 8.2 × 10 ⁻⁴ , C ₁ = −8.2 × 10 ⁻⁴ ( B) [Data 1 of low-pass filter corresponding to FIG. 7] 0 <r ≦ 2.2, C ₀ = 0, C ₁ = 6.21340780482349713 × 10 ^-4 C ₂ = −1.11831431665536618 × 10 ^-3 , C ₃ = 1.72326254774251052 × 10 ^{− 2} C ₄ = -7.4947015797950739 × 10 ^-2 , C ₅ = 0.15375365501776678 C ₆ = -0.17470277510871939, C ₇ = 0.11630783747263494 C ₈ = -4.52714464390924533 × 10 ^-2 , C ₉ = 9.56245557825196896 × 10 ^-3 C ₁₀ = -8.481108546 ^-4 (C) [Low-pass filter data 2 corresponding to FIG. 7] 0 <r ≦ 1.9, C ₀ = 0, C ₁ = 6.98116188060607758 × 10 ^-4 C ₂ = 3.176776369268199085 × 10 ^-5 , C ₃ = -2.53997915527410557 × 10 ^-4 C ₄ = 1.69910819567360878 × 10 ^-3 , C ₅ = -1.00618592831366177 × 10 ^-2 C ₆ = 2.59231882276316659 × 10 ^-2 , C ₇ = -3.1762710880 5453862 × 10 ^-2 C ₈ = 1.9706279862760988 × 10 ^-2 , C ₉ = -6.01906185353107658 × 10 ^-3 C ₁₀ = 7.21487026929249018 × 10 ^-4 (D) [Low-pass filter data 3 corresponding to FIG. 7] 0 <r ≦ _{1.9, C 0 = 0, C} 1 = 5.98410923155050407 × 10 -4 C 2 = 2.67892063030759041 × 10 -5, C 3 = -2.15027456191388931 × 10 -4 C 4 = 1.4457242736981412 × 10 -3, C 5 = -8.60047940280224323 × 10 - ³ C ₆ = 2.21876770919958392 × 10 ^-2 , C ₇ = -2.7198850592359095 × 10 ^-2 C ₈ = 1.6878252736166768 × 10 ^-2 , C ₉ = -5.15576206337112989 × 10 ^-3 C ₁₀ = 6.180304607001828 × 10 ^-4 (E) [Figure] Lowpass filter data 4 corresponding to 7] 0 <r ≦ 2.2, C ₀ = 0, C ₁ = 5.99973411495760274 × 10 ^-4 C ₂ = 7.24078912223179768 × 10 ^-7 , C ₃ = -9.50298609445933418 × 10 ^-6 C ₄ = 1.28239522689032291 × 10 ^-4 , C ₅ = -1.09084425624676046 × 10 ^-3 C ₆ = 3.41265021400184634 × 10 ^-3 , C ₇ = -4.60857251099639799 × 10 ^-3 C ₈ = 2.9131337750895013 × 10 ^-3 , C ₉ = -8.65112114233603337 × 10 ^-4 C ₁₀ = 9.80516170458266827 × 10 ^-5 (F) [Low-pass filter data 5 corresponding to FIG. 7] 0 <r ≦ 2.2, C ₀ = 0, C ₁ = -6.21386025668877481 × 10 ^-4 C ₂ = 1.11903042234138932 × 10 ^-3 , C ₃ = -1.72370859561942237 × 10 ^-2 C ₄ = 7.49614215496832575 × 10 ^-2 , C ₅ = -0.153780579745271 C ₆ = 0.17473340572566055, C ₇ = −0.1163293334441247 C ₈ = 4.52805238737387303 × 10 ⁻² , C ₉ = −9.56456858123039266 × 10 ⁻³ C ₁₀ = 8.48319025793002385 × 10 ⁻⁴ (G) [Low-pass filter data 6 corresponding to FIG. 7] 0 <r ≦ 1.9, _{C 0 = 0, C 1 =} 7.29467950111891425 × 10 -4 C 2 = -7.20406256342096506 × 10 -3, C 3 = 0.11227931692366647 C 4 = -0.48029263512608256, C 5 = 0.97855628529132503 C 6 = -1.1214666672199971, C 7 = 0.76512149197889504 C 8 = _{-0.30918978501733992, C 9 = 6.84811578855889364 × 10} -2 ^{_{10 = -6.41739833484116645 × 10 -3 (H}} ) [ data of the low-pass filter corresponding to FIG. 6] 0 <r ≦ 1, A 0 = 0, A 1 = 0.5 × 10 -3, B 1 = 0.45 × 10 -3 1 <r ≦ 2, A ₀ = z (when x ² + y ² = 1), A ₁ = −0.5 × 10 ⁻³ B ₁ = −0.45 × 10 ⁻³ (I) [Low-pass filter corresponding to FIG. 0 <r ≦ 2.2, C ₀ = 0, C ₁ = 1.18202302586474311 × 10 ^-3 C ₂ = 1.02432495935523004 × 10 ^-3 , C ₃ = -6.20002963466144939 × 10 ^-3 C ₄ = 1.52073827862137124 × 10 ^-2 , C ₅ = -1.88737777155327342 × 10 ^-2 C ₆ = 1.30665516375023426 × 10 ^-2 , C ₇ = -5.32356937854670548 × 10 ^-3 C ₈ = 1.27413320587890262 × 10 ^-3 , C ₉ = -1.66204863233887403 × 10 ^-4 C ₁₀ = 9.1386676289853178 × 10 ^{-6 In the} above data, r ₁ , r ₂ , ... are the curvature radii of each surface,
d _1, d _2, ··· surface separation, n _1, n _2, ··· is a refractive index of a medium such as a lens, ν _1, ν _2, ··· is also Abbe number. The r ₅ is a refractive surface derivative contains a discontinuous point (or line), r ₁ ~r ₃ is a data of the exit end of the endoscope.

The pixel pitch of the image sensor of the above embodiment is Px
= 0.0096 (horizontal direction) and Py = 0.0099 (vertical direction).

Nine types of refraction surfaces (A) to (I) are shown as examples of the refraction surface of the low-pass filter to be used, and a good low-pass effect can be obtained in each case.

[0042]

According to the image pickup apparatus of the present invention, an inexpensive low-pass filter is used, and an optimum low-pass effect according to a system to be used can be obtained.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a second embodiment of the present invention.

FIG. 3 is a cross-sectional view of a low-pass filter used in the present invention.

FIG. 4 is a diagram showing an image formation state when the above low-pass filter is used.

FIG. 5 is a diagram showing an inclination angle near a discontinuous point of the low-pass filter.

FIG. 6 is a sectional view of another low-pass filter used in the present invention.

FIG. 7 is a sectional view of another low-pass filter used in Embodiment 1 of the present invention.

FIG. 8 is a diagram showing a top line of a crystal filter arranged in a television camera.

FIG. 9 is a diagram showing a top line of another crystal filter arranged in the television camera.

FIG. 10 is a diagram showing a top line of a low-pass filter used in the present invention.

FIG. 11 is a diagram illustrating an MTF at a wide end according to the first embodiment.

FIG. 12 is a diagram illustrating an MTF at a telephoto end according to the first embodiment.

FIG. 13 is a diagram of another low-pass filter used in the present invention.

FIG. 14 is a diagram of still another low-pass filter used in the present invention.

FIG. 15 is a diagram showing an image forming situation when a conventional wedge prism is used.

FIG. 16 is a diagram showing a spatial frequency characteristic when the above prism is used.

Continuation of front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G02B 27/46

Claims (2)

(57) [Claims] 1. A imaging apparatus including an optical system which samples the light intensity to form an image on the imaging surface of the imaging device for imaging, the anamorphic refractive surface represented by the following formula in the optical system An imaging apparatus, comprising: a point image separated by the refraction surface. _{z = A 0 + A 1 x} + B 1 y + A 2 x 2 + B 2 y 2 + A 3 x 3 + B 3 y 3 + ··· + A n x n + B n y n (A 1 ≠ 0 and B ₁ ≠ 0) However, z Is an optical axis direction, and x and y are directions perpendicular to the optical axis. 2. An optical element used in an image pickup apparatus including an optical system for forming an image on an image pickup surface of an image pickup element which samples light intensity and picks up an image, wherein the anamorphic refraction is represented by the following equation: An optical element having a surface and separating a point image by the refraction surface. _{z = A 0 + A 1 x} + B 1 y + A 2 x 2 + B 2 y 2 + A 3 x 3 + B 3 y 3 + ··· + A n x n + B n y n (A 1 ≠ 0 and B ₁ ≠ 0) However, z Is an optical axis direction, and x and y are directions perpendicular to the optical axis.