JPS588483B2 - optical low pass filter - Google Patents

Description

Detailed Description of the Invention For example, when configuring a color solid-state imaging device using a solid-state imaging body such as a charge-coupled device (CCD), one way to obtain a certain degree of color resolution is to A method is generally known in which the frequency characteristics are set such that the passbands of the two are different from each other.

That is, as shown in FIG. 1, the band of green light G that directly affects the resolution is set relatively wide, for example, about 1 to 2 MHz, as shown by curve 1.

On the other hand, the bands of blue light B and red light R, which do not directly affect resolution, are made narrower than that of green light G.

For example, as shown in curve 2, the passband is approximately half (0.5
~1. 0 MHz).

The present invention proposes an optical filter, that is, an optical low-pass filter, which is suitable for application to the optical system of this type of solid-state imaging device, which is configured such that the pass bands handling each color light are different. In particular, in the present invention, this optical filter is configured as a wavelength-dependent filter, so that the same filter can be provided with different filter characteristics.

The optical filter of the present invention includes a flat plate made of an optically active substance and two
It consists of two or more birefringent plates, and the flat plate is constructed by stacking these plates interposed between the birefringent plates.

The optical filter of the present invention will be explained below with reference to the drawings.

In the present invention, a birefringent plate made of quartz or the like is prepared.

It is generally known that this refracting plate can be used as a low-pass filter.

That is, in the case where the object 3 is imaged through the optical lens system 4 as shown in FIG.
On the other hand, the image 6B due to the extraordinary ray LE is obtained with a horizontal shift of ΔX.

Note that this birefringent plate 5 is used with its optical axis Q being tilted so as to be parallel to the horizontal direction of the imaging plane.

FIG. 3 is a general explanatory diagram showing birefringence development of the birefringence plate 5, and the names of each part will be added first.

However, the illustrated example is a case where the birefringent plate is cut in the direction of the thickness d of the birefringent plate.

Q is the optical axis specific to the birefringent plate, nl is the refractive index in the optical axis direction, n2 is the refractive index for the extraordinary ray LB in the direction orthogonal to the optical axis, and ΔX is the extraordinary ray LE and the ordinary ray L.

and α is the extraordinary ray L.

and the ordinary ray Lo, and β is the angle between the optical axis Q and the incident ray.

Here, the angle α between the light rays LO and LE is given by, and this tanα takes the maximum value when tanβ11.

In the case of crystal, the difference between nl and 02 is about 0.5%, so the angle α is maximum when β=45°.

Therefore, when a light beam is incident at an angle of 45° with respect to the optical axis Q, the maximum separation distance ΔX can be obtained with the minimum thickness d, so β is generally selected to be 45°.

In this example as well, β is selected to be 45°.

Referring again to FIG. 2, the fact that the object 3 is imaged optically shifted by ΔX means that when considering this electrically, as shown in FIG. This is equivalent to the relationship between the delayed output and the non-delayed output when the delay circuit 8 is provided between them.

In this example, the time τ corresponding to ΔX is selected as the delay time.

Here, the transfer function R(f) of the circuit shown in the figure is...
...(1) Represented by:

Therefore, the amplitude is cos2πf, but when considering the response of imaging in the optical system shown in Fig. 2, the response in the horizontal direction can be expressed as shown in Fig. 4. Ru.

Therefore, equation (2) can be illustrated as shown in FIG. 5.

And the zero point direction of R(u) is given by.

If you look at equations (2) and (3) and FIG. 5, it will be clear that this is indicated by a line.

), this birefringent plate can be used as a low-pass filter.

The present invention utilizes this characteristic, and the present invention uses at least two birefringent plates.

An example of the optical filter 10 is shown in FIG.

In this example, three optical plates 5A to 5C are used, and as described later, the two birefringent plates 5A and 5B are provided on both upper and lower surfaces of a flat plate 5C having optical rotation.

That is, the optical filter 10 is constructed by laminating layers.

By the way, optically active substances occur because there is a difference in refractive index for circularly polarized light that rotates in opposite directions.For example, in quartz crystal, when the plane perpendicular to the optical axis Q is By using a thin plate obtained by cutting, optical rotation can be imparted to the thin plate.

Next, let us find the optical rotation angle γ of the flat plate 5C with this optical rotation.

A linearly polarized light wave P is shown in FIG. 7A.

This polarized light wave P can be considered to be a composite wave of a right-handed circularly polarized light wave PR and a left-handed circularly polarized light wave PL. Therefore, the refractive index for the right-handed circularly polarized wave is Nr, and that for the left-handed circularly polarized wave is Nl. In this case, the angle γ formed by the composite wave P is ? It can be seen that the light rotates by γ.

That is, the angle of optical rotation γ is proportional to the reciprocal of the wavelength λ, and the angle of optical rotation γ is proportional to the reciprocal of the wavelength λ.
is proportional to the thickness d.

Therefore, each of the R-H primary color lights has a different angle of optical rotation.

If the wavelengths of R-B are λR~λB, respectively, λB〈λG〈
Since the relationship λ holds true, if G is used as a reference for the angle of optical rotation γ, then the relationship between the angles of rotation of the husband and husband is shown in Figure 1B, where B shows the maximum optical rotation, and conversely, R shows the minimum angle of optical rotation. shows.

R, ~B, and G are respective angles of optical rotation.

The optical filter 10 will be explained based on the optical characteristics of the flat plate 5C as described above, but the optical axes Qa and Qb of the pair of birefringent plates 5A and 5B arranged above and below are as follows. selected based on the relationship.

When the optical axis Qa of the birefringent plate 5A currently located above is selected to be parallel to the X axis as shown in FIG.
The optical axis Qb of is selected to be parallel to Qa (θ10).

It should be noted that FIG. 8 shows a plan view of the birefringent plate, so that the light rays travel from the page toward the reader, and are called extraordinary rays L.

contains the vector E shown in the figure in the plane perpendicular to the plane of the paper.In other words, it has an electrical displacement in the plane that has the plane of the paper and the cutting line in the X-axis direction, so it becomes a linearly polarized light component in the X-axis direction. .

Similarly, the ordinary ray Lo is expressed as a linearly polarized component in the direction perpendicular thereto, that is, in the y-axis direction.

Next, the optical characteristics of this optical filter 10 will be explained with reference to FIG. 10 and subsequent figures.

First, considering the G component, the first birefringent plate 5
The G component incident on A is separated into an ordinary ray LOG and an extraordinary ray LEG.

Let that distance be Wl. These light rays LOGt LEG are rotated by a predetermined angle in a desired direction due to the presence of the flat plate 5C.

In this example, the flat plate 5 is rotated so that the green light G rotates by 270°.
The thickness d3 of C is selected (see FIG. 9).

Therefore, the light beams LOG and LEG rotate by the amount indicated by ■ in FIG. 10A.

These rotating lights LOG and LEG are each connected to the second birefringent plate 5B.
is the incident ray of light.

Therefore, the light ray LEG shown by ■ in FIG. 10A becomes an ordinary ray in the second birefringent plate 5B, and conversely, the ordinary ray LOG becomes an extraordinary ray.

Therefore, the light beam LOG is separated by W2, that is, the two polarized light components shown in (■) in FIG. 10A are finally obtained.

Next, let's consider red light R and blue light B.

Since the first birefringence 5A has no wavelength dependence, it is separated into light rays LoB (LoR) and LEB (LER) similarly to the green light G (see I in FIG. 10B).

Both of these rays are incident on the flat plate 5C,
Since this refracting plate 5C has wavelength dependence as shown in FIG. 7B, the light of R and B is rotated in the directions indicated by black in FIG. 7B, respectively.

Light rays LQB (LQR), LB rotated by a predetermined angle
When H(LER) is divided into components in the X-axis and y-axis directions, components as shown by dotted lines (see B II' in the same figure) are obtained for each light ray, so these components are When the light passes through, a plurality of polarized light components are finally obtained as shown in (■) in FIG.

Since all of the finally obtained light rays have components in the X-axis direction, the responses RU(1(u), RUB(u), and RUB(u) of R to B in the X-axis direction are respectively It becomes as shown in the following formula.

Now, if we choose W1 = 2 W2, equations (4) and (5) become equations (6) and (7) shown below, respectively.

and the response RUB(u), RUR(Ll
) as shown in curves 11R and 11B.

As shown above, three flat plates are used, and one of them
A wavelength-dependent optical filter 10 can be obtained by composing the filter 10 with a material having optical rotation.

Therefore, the optical filter characteristics of G and the optical filter characteristics of R and B are obtained respectively differently, that is, the 11th
As shown in the figure, if the angle of rotation γ is selected to be 270° at θ-0, the R and B bands can be set to i of the G band, and the first
The desired optical filter shown in the figure can be obtained.

Therefore, it is extremely suitable for application to an optical system in a color solid-state imaging device using a solid-state imaging body such as the CCD mentioned at the beginning.

In addition, in the above-mentioned embodiment, one flat plate 5C having optical rotation is used.
Although this is a case where the optical filter 10 is constructed using two birefringent plates 5A and 5B, this type of filter 10 can also be constructed by adding one more birefringent plate.

FIG. 12A shows an example of an optical filter 10' having this configuration.

In this case, if the added birefringent plate is 5D, its optical axis Qd is used with its optical axis Qd rotated by a predetermined angle θ in the y-axis direction with respect to the other optical axis Qa (Qb) (Fig. 12). (See B).

However, this example shows an example in which G is optically rotated by 180°.

Further, in this example, the optical axis Qb of the second birefringent plate 5B is line-symmetrical with respect to the optical axis Qa and the incident light beam.

That is, the optical axis is set as shown in FIG. 12C.

Even in this case, the projection diagram of the optical axis Qb onto the incident plane is the first one.
As shown in FIG. 2B, it is made to be parallel to the X-axis direction like the optical axis Qa.

Setting the optical axis Qb as shown in FIG. 12C can be achieved, for example, by rotating the birefringent plate 5A by 180° about the axis of the incident light beam.

When the optical axis Qb is set in this way, as is clear from the figure, the shift of the extraordinary ray LE is in the opposite direction compared to the birefringent plate 5A, that is, in the right direction with respect to the ordinary ray LQ as seen in the drawing. .

The component relationships for explaining the optical characteristics of this optical filter 10' are as shown in FIGS. 13A to 13D.

Although the detailed explanation is omitted, this optical filter 10
The respective responses obtained by ′ are as follows.

However, W2 and W3 are the birefringent plates, the separation distance between 5B and 5D, and θ
is the rotation angle of the optical axis Qd. Therefore, if equations (8) and (9) are illustrated, the 14th
The filter characteristics shown in the figure can be obtained.

In this case, the advantage of rotating the optical axis Qd by θ is that the G characteristic curve 12
In the case of G (R, H, 12R, 12B, the subsequent response will be much smaller than the previous response, so
You can enter as few initial trap points as possible.

That is, when a CCD is used as an imaging device, input light information corresponding to an image of a subject to be imaged is obtained from the CCD as an output signal sampled for each picture element.

That is, the sampling frequency is now f.

In this case, by scanning each pixel in each horizontal section, the output video signal obtained in one horizontal section is the 15th
As shown in the figure, in addition to the modulation component SDO of the luminance signal SY, the sampling frequency f.

A sideband component SM (alternating current component) modulated is obtained.

However, the illustrated example shows only the fundamental wave.

In this case, upper and lower sidebands are generated in the AC component SM around the sampling frequency f, so if the modulation component SDO has a sufficient band to prevent resolution deterioration, the first
As shown in Figure 5, the high frequency component SDH of the modulation component SDO
A sideband component due to the sampling frequency f 2 overlaps in the middle.

In other words, the shaded portion is caused by aliasing distortion.

If the image is played back in this state, a flickering phenomenon will appear on the playback screen.

In order to prevent this image deterioration, the folded components should be made as small as possible.

Therefore, if the optical axis Qd is rotated by θ as in this example, the response after the first trap point corresponding to the sideband component becomes smaller, so this optical filter 10' is used in the optical system of this type of imaging device. If applied, the above-mentioned drawbacks can be effectively eliminated simply by processing the optical system.

Although quartz is used as the material constituting the refracting plate in the above example, any other refractive plate may be used as long as it has a birefringence phenomenon.

Further, the number of refracting plates used is not limited.

Of course, devices to which the optical filters 10 and 10' can be applied are not limited to imaging devices.

In addition to the above-mentioned crystal, optically active substances include substances having electro-optical properties, that is, the Kerr effect, such as KD.
P (the molecular formula of which is shown as KH2PO4) can also be used.

[Brief explanation of drawings]

Figure 1 is a diagram showing the band characteristics of an optical filter, Figures 2 and 3 are diagrams used to explain the birefringence plate, Figure 4 is an explanatory diagram when considering the birefringence phenomenon electrically, and Figure 5 is a diagram showing the band characteristics of an optical filter. The figure is a diagram of its characteristic curve, Figure G is a configuration diagram of an optical filter according to the present invention, Figure 7 is a diagram showing optical rotation, Figure 8 is a diagram showing the direction of electric field in a birefringent plate, Figure 9 and FIG. 10 is a diagram for explaining FIG. 6, FIG. 11 is a band characteristic curve diagram obtained by the filter of the present invention, and FIGS. 12 to 14 are explanatory diagrams of other examples of the present invention. , FIG. 15 is a frequency spectrum diagram of an imaging signal obtained by a solid-state imaging device. 10 is an optical filter, 5A, 5B and 5D are birefringent plates,
5C is a flat plate having optical rotation, Qa=Qd is the optical axis, θ is the rotation angle, Wl-w3 is the separation distance, LO and Lo' are ordinary rays, and LE and LE' are extraordinary rays.

Claims (1)

[Claims] 1 It has a flat plate having optical rotation and two or more birefringent plates, the flat plate is interposed between the birefringent plates and is laminated together, and the flat plate has an optical rotation angle with respect to the incident light. The angle of optical rotation for incident light of a specific wavelength, which depends on the wavelength of the incident light and whose pass band should be selected broadly, is determined by the angle of optical rotation for the incident light of a specific wavelength, which depends on the wavelength of the incident light and whose pass band should be selected broadly. What is claimed is: 1. An optical low-pass filter, characterized in that the filter is selected so as to coincide with the direction perpendicular thereto, and has a different pass band for each incident color light.