JP2007158825A - Image input device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To cope with the problems that if a lens size is reduced in order to deepen a depth of field when focusing is impossible, the number of pixels in an image pickup element constituting an ommatidium image imaged by one lens is reduced, and consequently, resolution of a single image obtained by reconstructing a compound eye image is deteriorated, and also, if an electrowetting lens having a focusing function is used, it is possible to focus on a plurality of subject distances, however, it is necessary to know a subject distance to be focused on in an image input device using a lens array and the image pickup element. <P>SOLUTION: The ommatidium image formed on the image pickup element 3 when the same subject 6a is viewed through two lenses 11a, 11b located at different positions becomes an image formed by images that are shifted by an amount δ from each other. It is possible to calculate a distance L from each lens to the subject if a focus distance f of each lens, a distance d between the lenses, and the shift amount δ are known. A curvature radius of each lens is controlled so as to make the L as a focusing distance. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image input device using an optical system and an image sensor such as an image reading device, an image recognition device such as iris authentication and face authentication, a digital camera, and a surveillance camera.

A device using a lens array is known as a conventional image input device with a reduced thickness. However, since the distance between the lens array and the image sensor is fixed, focus adjustment cannot be performed (for example, Patent Document 1, 2). If the focus is set to the hyperfocal distance, even if the lens does not have a focus adjustment function, the depth of field of the lens is used to focus from half the hyperfocal distance to infinity. Although an image can be shot, when shooting a subject closer than half the hyperfocal distance, the defocus aberration increases as the subject distance decreases, resulting in an out-of-focus image. In addition, if the lens diameter is reduced, it is possible to obtain an image in which the depth of field is widened and an in-focus subject is in focus. The number of pixels decreases, and the resolution of a single image obtained by reconstructing a compound eye image decreases.
Even if a microlens array having means for focusing on the plurality of subject distances is provided, the function cannot be used effectively unless the subject distance to be focused can be known. You may focus on various subject distances to find out where the subject is in focus, but it takes time to do so, and it is inaccurate because it is necessary to judge whether the subject is in focus or not. There is a fear.

Although there is a conventional technique of a lens array having a focus adjustment function (see, for example, Non-Patent Document 1), an external device for controlling a fluid used for deformation of the lens is required, and each lens has a fluid flow path. In this case, each lens shape, that is, the focal length cannot be controlled separately.
In addition, there is an electrowetting lens having a focus adjustment function. However, the electrowetting lens has a drawback that it cannot be applied to a bright optical system because it is difficult in principle to increase the diameter. In the electrowetting lens, the focus adjustment function is realized by changing the liquid shape by utilizing the change in wettability between the conductive liquid and the dielectric due to the application of voltage. When the diameter is increased, the ratio of the surface area to the volume decreases, so that the influence of the wettability on the factor governing the liquid shape is reduced, and the physical quantity related to the volume such as the weight of the liquid becomes dominant. For these reasons, it is difficult to apply an electrowetting lens to a bright optical system. However, each lens constituting the lens array is significantly smaller than a lens used in a conventional image input device such as a camera. Therefore, the advantage of the electrowetting lens that focus adjustment is possible can be fully utilized.

JP 2001-061109 A JP 2003-141529 A JP 2001-515539 A Nikolas Chronis et al., “Tunable liquid-filled microlens array integrated with microfluidic network”. OPTICS EXPRESS Vol. 11, No. 19, pp2370

  Provided is an image input device capable of acquiring a focused image from a thin and close subject to a subject at infinity.

According to the first aspect of the present invention, the lens array includes a lens array composed of two or more lenses, a light shielding means for preventing crosstalk of light rays between the lenses constituting the lens array, and the lens array. An image input apparatus comprising: an imaging unit for capturing an imaged image; and a first computing unit for reconstructing a single image from a compound eye image captured by the imaging unit. It has a function capable of focusing on at least a plurality of subject distances.
According to a second aspect of the present invention, in the image input device according to the first aspect, the lens array is an electrowetting lens.
According to a third aspect of the present invention, in the image input device according to the second aspect, the image input apparatus further comprises distance detection means for detecting a subject distance.
According to a fourth aspect of the present invention, in the image input device according to the third aspect, the distance detecting means is a second computing unit that obtains a subject distance based on parallax information of a compound eye image.

According to a fifth aspect of the present invention, in the image input device according to the fourth aspect, the calculation for detecting the parallax of the compound eye image by the second calculator is a cross-correlation calculation of the compound eye image. Features.
According to a sixth aspect of the present invention, in the image input device according to any one of the third to fifth aspects, the focus distance is changed by feeding back the output of the subject distance detection means to the shape of the electrowetting lens. It has the means to make it have, It is characterized by the above-mentioned.
According to a seventh aspect of the present invention, in the image input device according to any one of the second to fifth aspects, the lens array includes a plurality of lenses having the same in-focus distance as a single lens set. It is an aggregate of electrowetting lenses composed of a plurality of lens sets having different focal distances.
According to an eighth aspect of the present invention, in the image input device according to the seventh aspect, depths of field by the plurality of lens sets are at least continuous with each other.
According to a ninth aspect of the present invention, in the image input device according to the seventh or eighth aspect, a single-eye image for reconstructing the single image from the compound eye image for each of the plurality of lens sets. It further has a single eye image selection means for selecting.

According to a tenth aspect of the present invention, in the image input device according to the sixth or ninth aspect, the in-focus portion extracting unit is used to focus on a plurality of single images reconstructed for each lens having a different in-focus distance. And an arithmetic unit for extracting and synthesizing the in-focus portion.
According to an eleventh aspect of the present invention, in the image input apparatus according to the tenth aspect, the focused part extracting means is an arithmetic unit that extracts a part having a high contrast as a focused part.
According to a twelfth aspect of the present invention, in the image input device according to any one of the second to eleventh aspects, the image input device further includes a sealing member for preventing contact between the electrowetting lenses. To do.
According to a thirteenth aspect of the present invention, in the image input device according to the first aspect, the lens array includes a plurality of lenses having the same focus distance by having the same thickness, and a lens set. It is a lens array composed of a plurality of lens sets having different focal distances.

According to a fourteenth aspect of the present invention, in the image input device according to the first aspect, the lens array includes a plurality of lenses having the same focal distance by having the same radius of curvature as one lens set, It is a lens array composed of a plurality of lens sets having different in-focus distances.
According to a fifteenth aspect of the present invention, in the image input device according to the thirteenth or fourteenth aspect, depths of field by the plurality of lens sets are at least continuous with each other.
According to a sixteenth aspect of the present invention, in the image input device according to any one of the thirteenth to fifteenth aspects, the image input device further includes a distance detecting means for detecting a subject distance.
According to a seventeenth aspect of the present invention, in the image input device according to the sixteenth aspect, the distance detecting means is a second computing unit that obtains a subject distance based on parallax information of a compound eye image.
According to an eighteenth aspect of the present invention, in the image input device according to the seventeenth aspect, the calculation for detecting the parallax of the compound eye image by the second calculator is a cross-correlation calculation of the compound eye image. Features.

According to a nineteenth aspect of the present invention, in the image input device according to any one of the sixteenth to eighteenth aspects, the single image is reconstructed from the compound eye image based on the output of the distance detecting means. It further has single-eye image selection means for selecting a single-eye image for performing.
According to a twentieth aspect of the present invention, in the image input device according to any one of the thirteenth to nineteenth aspects, the in-focus portion extracting unit from a plurality of single images reconstructed for each of the plurality of lens sets. Thus, it further has an arithmetic unit that extracts and synthesizes the in-focus portion in focus.
According to a twenty-first aspect of the present invention, in the image input apparatus according to the twentieth aspect, the in-focus portion extracting means is an arithmetic unit that extracts a high-contrast portion as the in-focus portion.

  According to the present invention, since the lens array is provided with means for focusing on a plurality of subject distances, an image input device capable of acquiring a focused image from a thin and close subject to a subject at infinity is provided. Can be provided.

FIG. 1 is a schematic diagram for explaining the configuration of the present invention. FIG. 4A shows a case where the focusing distance of the lens is long, and FIG. 4B shows a case where the focusing distance of the lens is short.
In the figure, reference numeral 1 denotes a lens array, 2 denotes a sealing member, 3 denotes an imaging element, 4 denotes a light shielding member, 5 denotes an information processing circuit, 6 denotes a subject, and 7 denotes a control circuit for adjusting the lens focus.
The lens array 1 is composed of two or more electrowetting lenses arranged on a substantially plane. The sealing member 2 is a member provided to prevent contact between the electrowetting lenses. The image sensor 3 is installed at a substantially image forming position of a compound eye image formed by the lens array. The light shielding member 4 is provided in order to prevent light beam crosstalk between the lenses constituting the lens array. Of the information processing circuit 5, 5 a is a computing unit that detects parallax information of the compound eye image, and 5 b is a distance detection unit that uses the parallax information to detect the distance from the image input device to the subject. (Second computing unit) 5c is a computing unit (first computing unit) for reconstructing a single image from the compound eye image.
Since the subject 6 generally has a depth, in FIG. 5A, when the electrowetting lens is in the initial state, the subject 6a is close to the subject depth within the depth of field. It is represented by two planes of the subject 6b on the side. However, the broken line representing both represents the position where the subject exists, and does not represent that the subject 6b blocks the subject 6a.

  The subject 6a existing within the depth of field of the lens 1a set in the initial state is imaged by the respective electrowetting lenses 1a constituting the lens array, and is imaged by the imaging device 3. The light shielding member 4 suppresses the crosstalk at the position of the imaging element of the light beam passing through each lens. Thus, the image picked up by the image pickup device 3 becomes a compound eye image in which single-eye images picked up by the respective lenses are arranged. The single-eye images constituting the compound-eye image appear to be the same image at first glance, but each single-eye image becomes a slightly shifted image because of the parallax caused by the relative positional relationship between the lens and the subject. The “parallax” used in the present specification represents the shift amount (unit: length) of each single-eye image relative to the reference single-eye image in the recorded compound-eye image. If only a single-eye image is used as a recorded image by the image input device, it is not possible to capture the structure of a subject smaller than one pixel in the single-eye image, but if the parallax between single-eye images is used, The structure of a subject buried in one pixel can be reproduced. That is, by reconstructing a single image from a compound eye image with parallax, it is possible to obtain a reconstructed image with improved resolving power for a single-eye image (see Patent Document 1).

The lens focus adjustment circuit 7 applies a predetermined voltage to the electrodes of the electrowetting lens to change each lens to have a desired radius of curvature, and sets the lens to a specific focus distance. Other focus distances can be selected by changing the voltage to other values. Here, the in-focus distance is a subject distance that is correctly focused on the imaging surface when the lens specifications and the imaging distance are fixed.
In FIG. 1 (a), the light emitted from one point is imaged on the subject 6b on the side far from the depth of field of the lens 1a set in the initial state, as indicated by the broken light path. The image does not converge to one point on the element 3 and a so-called blurred image is obtained. Therefore, as in the lens 1b shown in FIG. 1B, the subject 6b cannot be focused unless the radius of curvature of the lens surface is changed to reduce the focusing distance. How to change the in-focus distance can be calculated by knowing the parallax.

FIG. 2 is a diagram illustrating an example of a cross-correlation function.
The computing unit 5a is for obtaining the parallax between each single-eye image in the recorded compound-eye image and, as described above, for increasing the resolving power for the single-eye image when reconstructing a single image from the compound-eye image. At the same time, as shown below, it also acts to obtain the distance from the parallax to the subject.
In order to obtain the parallax of each single-eye image from the compound-eye image, two single-eye images are extracted from the compound-eye image and their cross-correlation function is obtained. This cross-correlation function is a two-variable function and shows a peak at a certain coordinate. When the parallax is zero, the peak coordinate position is in the center, and the peak coordinate is away from the center as the parallax is increased. Using the distance and direction between the peak coordinates and the center, it is possible to quantitatively detect the relative positions of the two single-eye images, that is, parallax information.

FIG. 3 is a diagram showing an example of a compound eye image.
A compound eye image is an image in which reduced images of a subject are formed by the number equal to the number of microlenses. As shown in FIG. 6 to be described later, since the subject distance is much larger than the focal length of each lens, even if the subject has a depth and is composed of objects or persons having a plurality of subject distances, The fields of view are almost identical to each other, and the difference between them cannot be seen at first glance. Strictly speaking, however, the lenses at different positions see slightly different subject areas, so the formed single-eye image is slightly shifted due to the difference in viewing position, so-called parallax. . As a method for determining the amount of deviation between two substantially identical images, a method for obtaining a cross-correlation function as described above is known.

  Of the single-eye images constituting the compound-eye image, a reference one (for example, 50 in the figure) is set, and the cross-correlation function between the single-eye image and the other single-eye image is obtained. By obtaining a cross-correlation with the reference single-eye image for all single-eye images other than the single-eye image, the image shift amount of all single-eye images with respect to the reference single-eye image is obtained in units of pixels of the image sensor The parallax can be calculated by multiplying the image shift amount (unit pixel) by the length per pixel of the image sensor. Although it is possible to use an image that is out of focus in order to obtain parallax, it is desirable that the degree of defocus is as small as possible. For this reason, it is desirable to set the initial state of the electrowetting lens at a focus distance that is in focus with respect to the depth of field as deep as possible, that is, pan focus. Or, in addition to this, by providing a plurality of electrowetting lenses having an in-focus distance shorter than that of the pan focus, a plurality of objects that are in focus with respect to a subject that is closer than half the hyperfocal distance can be obtained. A single-eye image can be acquired, and parallax can be easily detected. In this case, the depth of field with the lens with the shorter focus distance is usually relatively shallow, but it overlaps with the depth of field (pan focus range) with the focus distance of the other lens. Therefore, it is desirable to set the in-focus distance so that it is at least continuous.

FIG. 4 is a diagram for explaining the triangulation method.
In the figure, 11 is a lens, d is a lens center distance, L is a subject distance, O is an intersection of a lens optical axis and an image sensor, P is a specific subject, P ′ is an imaging point of a specific subject, and δ is a parallax. Show.
FIG. 5 is a diagram showing a flow of obtaining a single image from a compound eye image.
In the figure, the symbol S indicates a flow step.
The computing unit 5b is for detecting the distance from the image input device to the subject using the parallax information. As a method for obtaining the subject distance, for example, a triangulation method using the principle of triangulation is known. The method for obtaining the subject distance by triangulation will be described with reference to FIG. Of the light emitted from the point P where the optical axis of the lens 11 a intersects the subject 6, the light that has passed through the lens 11 a forms an image at a point Oa where the optical axis of the lens 11 a and the image sensor 3 intersect. On the other hand, the light passing through the lens 11b forms an image at a point P ′ on the image sensor. If the point where the optical axis of the lens 11b and the image sensor 3 intersect is Ob, the parallax is P'Ob. The subject distance L is determined from the proportional relationship of the triangle by the lens center interval d and the lens array-imaging device distance (lens focal length) f known at the time of designing the optical system, and the parallax P′Ob = δ obtained by the calculator 5a. It can be obtained by the following formula.
L = f × d / δ (1)
In the present invention, this δ is calculated using the cross-correlation function described above.

When a compound eye image is obtained (S1), a reference single image 50 is set as shown in FIG. 3 (S2). The parallax δ of each single-eye image with respect to this reference single-eye image is detected (S3). The subject distance L is calculated from this parallax (S4).
The subject distance L is fed back from the distance detection means 5b to the control circuit 7 of the electrowetting lens 1 (see FIG. 1B), and the subject is focused by changing the shape of the electrowetting lens 1. Can do. An autofocus function can be added by automatically performing these parallax detection (S3) to lens shape change (S5). For example, when focusing on the subject 6b of FIG. 1B closer to the lens from the state of FIG. 1A focused on the subject 6a, based on the subject distance calculated based on the parallax. As in 1b, the shape of the electrowetting lens is changed to shorten the focal length of the lens, and a compound eye image in focus on the subject 6b can be captured (S6). By reconstructing this compound eye image into a single image by the calculator 5c (S7), it is possible to obtain a high-resolution image focused on the close subject 6b.
According to this method, a simple configuration for detecting a subject distance can be provided without requiring extra components.

FIG. 6 is a diagram showing another embodiment of the present invention.
In the figure, in order to show that the subject distance is very large compared to the lens focal length, the distance on the subject side is reduced by dividing it with a broken line. The same applies to the following drawings.
FIG. 7 is a diagram showing a flow for obtaining a single image corresponding to the configuration of FIG.
As shown in FIG. 6, the shape of each electrowetting lens is adjusted by controlling the shape so that one lens 1a is focused on a subject 6a at a distant position, and another lens 1b is focused on a subject 6b close to the lens. By making the focal lengths different, it is possible to pick up an image focused on a subject far from a close subject in one shot. When a plurality of lenses 1a and 1b are used, a group of lenses having the same focusing distance is called a lens set. In this example, there are two sets of a lens set including the lens 1a and a lens set including the lens 1b.
For example, in an image input device using a lens array in which electrowetting lenses are arranged in a 4 × 4 grid, half of the eight lenses are set to the focusing distance, and the other half of the eight are pan-focused. In this way, there are two types of images: a compound eye image composed of eight single-eye images capable of focusing on a close subject and a compound eye image composed of eight single-eye images in pan focus. A compound eye image is obtained. Each of these compound eye images is reconstructed using parallax by the calculator 5c, so that a high-resolution single image focused on a close subject and a high-resolution single image of pan focus are shot in one shot. Can be obtained at It is assumed that a memory or the like for storing each single image is provided as necessary.
As described above, it is desirable to set the respective in-focus distances so that the depths of field of the two lens sets at this time are at least continuous.
Also in this configuration, the focusing distance of either one of the lens sets may be changed based on the subject distance calculated based on the parallax. In that case, a higher resolution single image can be reconstructed.

The above processing flow is shown in FIG. Compared to the case where all the electrowetting lenses that make up the lens array have the same focal length, if a photograph is taken using a lens array composed of electrowetting lenses that are controlled to have different focal lengths, the reconstructed single The resolution of one image is reduced. However, it is possible to obtain a high-resolution single image by increasing the number of pixels per lens using an image sensor having a small pixel size.
Although the two lens sets can be arranged separately from each other, the individual lenses of each lens set are arranged alternately so that a larger parallax can be obtained in order to improve the accuracy of subject distance detection. It is preferable that the lenses of each lens set are arranged so as to be substantially evenly distributed over the entire imaging region. This concept can be applied in the same way as the number of lens sets increases.

FIG. 8 is a view showing still another embodiment of the present invention.
In the figure, reference numeral 10 denotes a multifocal lens array, and 20 denotes a subject.
FIG. 9 is a diagram showing a flow for obtaining a single image corresponding to the configuration of FIG.
The present embodiment is an image input device in which the lens array is composed of lenses having different thicknesses as a function that allows the lens array to focus on a plurality of subject distances.
A method of focusing the lens array with respect to a plurality of subject distances will be described with reference to FIG. When the lens array 10 configured by arranging plano-convex lenses having different thicknesses as shown in the figure is used, if the curvature of the convex surface of each lens is the same, the greater the thickness, the shorter the focal length. When a lens with the same focal length is collectively referred to as a lens set. In the figure, there are three lens sets.
When the distance between the lens and the image sensor is fixed, the lenses with the larger lens thickness, such as the lenses 10a, 10b, and 10c, focus on the subjects 20a, 20b, and 20c, respectively. Can burn. The compound eye image picked up by the image input apparatus having such a configuration is composed of single-eye images partially focused on different subject distances.
When the subject distance is calculated by the computing unit 5b, the computing unit 5d determines a set of lenses having the in-focus distance closest to the subject distance. A set of single-eye images corresponding to the determined lens is selected and reconstructed by the calculator 5c using parallax, thereby automatically obtaining a high-resolution single image in which the subject is in focus. be able to.
As described above, it is desirable that the depths of field of the three lens sets are at least continuous.

FIG. 10 is a flowchart showing another method for obtaining a single image.
As described above, the subject is not always at a single distance. That is, when the subject has a depth, if each individual image is viewed microscopically, a portion that is in focus and a portion that does not match are mixed.
Therefore, instead of selecting and reconstructing only a single eye image set corresponding to a lens having a certain focus distance, a single image is obtained from each single eye image set corresponding to a lens having the same focus distance. An all-in-focus image can be obtained by acquiring, extracting a portion in focus from the plurality of single images, and synthesizing them. When combining, it is more desirable to combine the respective parts by adjusting the difference in magnification caused by the difference in focusing distance. Further, as a means for extracting a focused part in a single image, a focused part detection method using contrast may be used. That is, an in-focus portion extraction unit is provided, and an arithmetic unit that extracts a high-contrast portion as the in-focus portion is provided.
This method can be applied not only to the embodiment of FIG. 8, but also to the embodiment shown in FIG. In the configuration shown in FIG. 1, the in-focus portion is extracted from both the single image obtained by focusing on the distance of the subject 6a and the single image obtained by focusing on the distance of the subject 6b. By doing so, it is possible to obtain an omnifocal image. Furthermore, the present invention can be applied to the following embodiments.

FIG. 11 is a view showing still another embodiment of the present invention.
In the figure, reference numeral 30 denotes a lens, and 40 denotes a subject.
The lens array in which plano-convex lenses having different curvatures shown in the figure are arranged side by side has the same effect as the case of using the lens array having lenses having different thicknesses shown in FIG. The smaller the radius of curvature of the convex surface, the shorter the focal length. Therefore, as in the case of the lens array composed of lenses having different thicknesses, when the distance between the lens and the image sensor is constant, the lens sets 30a, 30b, and 30c are respectively applied to the subjects 40a, 40b, and 40c. A lens having a smaller radius of curvature of the lens can focus on a closer subject, such as focusing. The procedure for obtaining a high-resolution single image is the same as that shown in FIG.
As described above, it is desirable that the depths of field of the three lens sets are at least continuous.
In the present embodiment and the embodiment shown in FIG. 8, the lens array is described as using a refractive lens. However, even if a diffractive lens array such as a Fresnel lens is used, the same effect can be obtained. Obtainable.

It is a figure for demonstrating the structure of this invention. It is a figure showing an example of a cross correlation function. It is a figure which shows an example of a compound eye image. It is a figure for demonstrating a triangulation method. It is a figure which shows the flow which obtains a single image from a compound eye image. It is a figure which shows other embodiment of this invention. It is a figure which shows the flow for obtaining the single image corresponding to the structure of FIG. It is a figure which shows other embodiment of this invention. It is a figure which shows the flow for obtaining the single image corresponding to the structure of FIG. It is a flowchart which shows how to obtain | require another single image. It is a figure which shows other embodiment of this invention.

Explanation of symbols

1 Lens Array 3 Image Sensor 5 Information Processing Circuit 6 Subject 7 Control Circuit

Claims (21)

  A lens array composed of two or more lenses, a light shielding means for preventing crosstalk of light rays between the lenses constituting the lens array, and an image formed by the lens array An image input apparatus comprising: an imaging unit; and a first computing unit for reconstructing a single image from a compound eye image captured by the imaging unit, wherein the lens array matches at least a plurality of subject distances. An image input device having a function capable of being focused.   The image input device according to claim 1, wherein the lens array is an electrowetting lens.   The image input apparatus according to claim 2, further comprising a distance detection unit that detects a subject distance.   The image input device according to claim 3, wherein the distance detection unit is a second computing unit that obtains a subject distance based on parallax information of a compound eye image.   5. The image input device according to claim 4, wherein the calculation for detecting the parallax of the compound eye image by the second calculator is a cross-correlation calculation of the compound eye image.   6. The image input device according to claim 3, further comprising means for feeding back the output of the subject distance detection means to the shape of an electrowetting lens to change the focusing distance. An image input device.   6. The image input device according to claim 2, wherein the lens array includes a plurality of lenses having the same in-focus distance as one lens set, and a plurality of lens sets having different in-focus distances. An image input device comprising an assembly of configured electrowetting lenses.   8. The image input device according to claim 7, wherein depths of field by the plurality of lens sets are at least continuous with each other.   9. The image input device according to claim 7 or 8, further comprising: single-eye image selection means for selecting a single-eye image for reconstructing the single image from the compound-eye image for each of the plurality of lens sets. An image input device further comprising:   10. The image input device according to claim 6, wherein an in-focus portion that is in focus is extracted from a plurality of single images reconstructed for each lens having different in-focus distances, and is synthesized. An image input apparatus further comprising an arithmetic unit for performing the operation.   The image input apparatus according to claim 10, wherein the focused part extracting unit is an arithmetic unit that extracts a part having a high contrast as a focused part.   12. The image input device according to claim 2, further comprising a sealing member for preventing contact between the electrowetting lenses.   2. The image input device according to claim 1, wherein the lens array includes a plurality of lenses having the same focus distance by having the same thickness, and a plurality of lens sets having different focus distances. An image input device comprising a lens array configured.   2. The image input device according to claim 1, wherein the lens array includes a plurality of lenses having the same focal distance by having the same radius of curvature as one lens set, and a plurality of lens sets having different focal distances. An image input device comprising: a lens array comprising:   15. The image input device according to claim 13, wherein depths of field by the plurality of lens sets are at least continuous with each other.   16. The image input device according to claim 13, further comprising distance detection means for detecting a subject distance.   17. The image input device according to claim 16, wherein the distance detecting means is a second computing unit that obtains a subject distance based on parallax information of a compound eye image.   18. The image input device according to claim 17, wherein the calculation for detecting the parallax of the compound eye image by the second calculator is a cross-correlation calculation of the compound eye image.   19. The image input device according to claim 16, wherein a single-eye image is selected for reconstructing the single image from the compound-eye image based on the output of the distance detection unit. An image input apparatus further comprising a single-eye image selecting unit for performing the following.   21. The image input device according to claim 19, wherein the in-focus portion is in focus by a in-focus portion extraction unit from a plurality of single images reconstructed for each of the plurality of lens sets. An image input device further comprising an arithmetic unit that extracts and synthesizes.   21. The image input apparatus according to claim 20, wherein the focused part extracting means is an arithmetic unit that extracts a part having high contrast as a focused part.

Images