JP4758773B2 - 3D image scanner - Google Patents

Description

  The present invention relates to an image scanner for obtaining an orthographic projection image of a three-dimensional object.

  In Patent Document 1, an imaging device including a telecentric lens is moved in the imaging direction with respect to a three-dimensional object, and the three-dimensional object is repeatedly photographed by the imaging device, and a large number of images obtained thereby are captured. By analyzing the transition of the gray level in the direction, the in-focus position is detected in the images, and the pixel data of the in-focus position is collected, and the three-dimensional object in which all the pixels are in focus A three-dimensional object recognition device that synthesizes the entire image of the above is disclosed.

JP 2003-207320 A

  According to the three-dimensional recognition apparatus disclosed in Patent Document 1, the size of an object from which the entire image can be acquired is limited to the size of the field of view of the imaging device. That is, with this three-dimensional recognition device, it is not possible to obtain an entire image of a three-dimensional object larger than the field of view of the imaging device. Further, Patent Document 1 has no disclosure about a method of illuminating a three-dimensional object. However, when the subject is a three-dimensional object, a shadow generated on the surface of the three-dimensional object by illumination may adversely affect the image of the three-dimensional object.

  Accordingly, an object of the present invention is to enable an orthographic projection image of an entire three-dimensional object larger than the field of view of an imaging device to be obtained in a three-dimensional object image scanner.

  Another object of the present invention is to reduce an adverse effect on an orthographic projection image of a three-dimensional object due to a shadow caused by illumination in a three-dimensional object image scanner.

  Another object of the present invention is to obtain an entire image of a subject with high geometric accuracy in a three-dimensional object image scanner.

  A three-dimensional object image scanner according to one aspect of the present invention is an image scanner for obtaining an orthographic projection image of a three-dimensional object, and is located at a position away from the image sensor and the image sensor and the image sensor in the observation direction. A telecentric imaging system that is arranged between the target plane area and forms an image of the target plane area on the image sensor; and the target plane area in a three-dimensional direction in a subject space in which a subject is arranged A moving means for moving and irradiating the target plane area with illumination light, so that a distribution range of incident angles of the illumination light to the target plane area is constant at every point of the target plane area; Illumination means that regulates the direction of the illumination light and the target plane area move in the subject space During this time, image data sequentially output from the image sensor is input, pixel data in focus on the subject is detected from the input image data, and the detected pixel data is collected. Image processing means for synthesizing an orthographic projection image of the subject.

  In a preferred embodiment, the target plane area is a linear area. The illumination unit includes at least one linear light source arranged in parallel with the target plane area, and the linear light source maintains a certain positional relationship with the target plane area. It moves with the plane area. In addition, the length of the linear light source is such that the angle formed by two line-of-sight lines connecting both ends of the linear light source and both ends of the target planar region is equal to or greater than a predetermined angle. It is set longer than the length. Further, the linear light source has a light direction restricting means for restricting an emission direction of the irradiation light along a plane including the two line-of-sight lines to a range of the predetermined angle. Thereby, the incident light quantity of the illumination light is constant at every point in the target plane area, and the distribution range of the incident angle of the illumination light along the plane is constant within the range of the predetermined angle. It has become.

  In a preferred embodiment, the illuminating means has at least two light sources arranged at positions on both sides of the optical axis from the telecentric imaging system toward the target plane region. When the target plane area is viewed from the telecentric imaging system, one light source irradiates the target plane area almost from the front, and the other light source irradiates the target plane area from an oblique direction. Yes. Furthermore, the output light amounts of the at least two light sources can be variably set independently.

  In a preferred embodiment, the target plane region is a linear region having a predetermined length parallel to the Y direction perpendicular to the observation direction. The moving means divides the subject space into a plurality of layers having different positions in the observation direction, and each of the plurality of layers has a different position in the Y direction and a width in the Y direction. Is divided into a plurality of bands set equal to the length of the target plane area, and the target plane area is moved to the observation direction to place the target plane area at the position of each layer in the observation direction, The target plane area is arranged at the position of each band in the Y direction by moving the target plane area in the Y direction at the position of each layer, and the target plane area is positioned at the position of each band. The image data of each band is output from the image sensor by moving in the X direction perpendicular to the observation direction and the Y direction. The image processing unit generates the image data of the plurality of layers by collecting the image data of the plurality of bands output from the image sensor, and the generated image data of the plurality of layers is set as the X direction. By comparing each coordinate in the Y direction, pixel data in focus on the subject is extracted from the image data of the plurality of layers.

  In a preferred embodiment, the image processing means uses a white reference level detection means for detecting a white reference level of the image sensor, and the image data output from the image sensor using the detected white reference level. Shading correction means for performing shading correction on the image. A plurality of white reference signs having different brightness are provided, and when detecting the white reference level of the image sensor, a white reference sign having an appropriate brightness is selected according to the average brightness of the subject. Can be used. As a result, it is possible to obtain a gradation image faithful to the brightness (reflectance) of the subject without depending on the light amount or the light amount unevenness of the illumination device.

  A stereoscopic image scanner according to another aspect of the present invention is disposed between an image sensor and the image sensor and a target plane area remote from the image sensor, and images an image of the target plane area on the image sensor. A telecentric imaging system for imaging, a moving means for moving the target plane region in a three-dimensional direction in a subject space in which a subject is arranged, and both sides of an optical axis from the telecentric imaging system to the target plane region And at least two light sources for irradiating illumination light to the target plane area disposed at a position and sequentially output from the image sensor while the target plane area is moving in the subject space. Image data is input, and pixel data in focus on the subject is selected from the input image data. Out, and an image processing means for synthesizing the orthogonal projection image of the object attracted detected the pixel data. Then, the image processing means turns on only one of the at least two light sources to obtain and store a first sub-orthogonal projection image, and then turns on only another light source to turn on the second light source. A sub orthographic projection image is obtained and stored, and then the first and second sub orthographic projection images are integrated to finally obtain the orthographic projection image.

  According to this three-dimensional object image scanner, even when the illumination unevenness patterns of a plurality of light sources are extremely different, the difference in the illumination unevenness hardly appears in the finally obtained subject image, and the image quality is good.

  A stereoscopic image scanner according to still another aspect of the present invention is disposed between a linear image sensor and a linear target plane region that is separated from and parallel to the image sensor and the image sensor. A telecentric imaging system for forming an image of the target plane area on the image sensor; a moving means for moving the target plane area in a three-dimensional direction in a subject space in which a subject is arranged; and the target plane area Input image data sequentially output from the image sensor while moving in the subject space, and detect pixel data focused on the subject from the input image data, Image processing means for collecting the detected pixel data and synthesizing an orthographic projection image of the subject.

  Since this three-dimensional object image scanner uses a linear, that is, one-dimensional image sensor (linear image sensor), the image data of the obtained target area is also one-dimensional image data. Of the image data, those in focus are connected to each other in the width direction and the length direction, and an orthographic projection image of the subject is synthesized. According to this three-dimensional object image scanner, a telecentric imaging system is used as compared with the case where a two-dimensional image sensor (area image sensor) is used (that is, the image data of the target area is two-dimensional). Image distortion due to the effects of optical characteristics (such as lens aberration) is less likely to occur in the width direction of the image data, and in particular, there is a correction to be applied to the image data in order to connect the image data accurately in the length direction. Since it is sufficient to correct shear deformation as specifically described in an embodiment described later, it is much easier to obtain a subject image with high geometric accuracy.

  According to the present invention, the object is scanned while moving the tertiary direction in the target plane area, which is an object area that can be simultaneously photographed by the image sensor. Therefore, in the image scanner for a three-dimensional object, the entire three-dimensional object larger than the target is scanned. An orthographic projection image can be obtained.

  Also, according to one aspect of the present invention, the direction of the illumination light is regulated so that the distribution range of the incident angle of the illumination light to the target plane region is constant at every point in the target plane region. The adverse effect on the orthographic projection image of the three-dimensional object due to the caused shadow is reduced.

  In addition, according to another aspect of the present invention, even when the illumination unevenness patterns of a plurality of light sources are different, the influence of the difference between the illumination unevenness patterns can be made difficult to appear in the subject image.

  According to another aspect of the present invention, it is easy to obtain a subject image with high geometric accuracy.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 shows the overall configuration of a three-dimensional object image scanner according to an embodiment of the present invention.

  As shown in FIG. 1, a table 14 for placing a three-dimensional object (not shown) is installed on a platform 12 installed on the floor. The space in which the subject above the table 14 is arranged is hereinafter referred to as “photographed space”. In the following description, an orthogonal three-dimensional coordinate system (left-handed system) having X, Y, and Z axes as shown in FIG. 1 is defined for the convenience of explaining the three-dimensional positional relationship. Here, the Y axis is the vertical direction (height direction), and the X axis and the Z axis are horizontal directions.

  A columnar X track 16 is laid parallel to the X axis from the table 14 on the platform 12 to the front. A columnar Y track 18 is mounted on the X track 16 in parallel to the Y axis. A columnar carriage 20 is attached to the Y track 18 in parallel to the Z axis. The Y track 18 incorporates an electric self-propelled device (not shown) and can reciprocate in a direction (lateral direction) parallel to the X axis within the length of the X track 16. The carriage 20 also incorporates an electric self-propelled device (not shown) and can reciprocate in the direction (height direction) parallel to the Y axis within the range of the length of the Y track 18. It can also reciprocate in the direction parallel to the Z axis (depth direction) within the length range. As a result, the carriage 20 can move three-dimensionally along the X, Y, and Z axes.

  A plate-like pedestal 22 is fixed to the carriage 20, and a scanning unit 24 and an illumination device 32 are fixed to the pedestal 22. The scanning unit 24 is an optical device for three-dimensionally scanning a subject (not shown) arranged in the subject space 15, and includes a telecentric imaging unit 26 arranged in parallel with the Z axis and a variable aperture. A unit 27 and an image sensor unit 28 are provided. The internal configuration and the function of the scanning unit 24 will be described later with reference to FIGS. 2 and 3, but briefly described as follows.

  The image sensor unit 28 incorporates a linear (linear) linear image sensor (reference number 46 in FIGS. 2 and 3) arranged in parallel with the Y axis. The telecentric imaging unit 26 has a built-in telecentric lens system (reference numeral 40 in FIGS. 2 and 3), and the optical axis 43 is parallel to the Z axis so that the positive direction (observation direction) of the Z axis is set. It is arranged facing. The telecentric imaging unit 26 has a linear (that is, linear) planar area (hereinafter referred to as a linear area) having a predetermined length parallel to the Y axis at a predetermined distance from the front end 26A along the optical axis 43 in the observation direction. 48 images (referred to as target plane regions) are formed on a linear image sensor in the image sensor unit 28. That is, the scanning unit 24 is in focus with the target plane area 48. The variable aperture unit 27 is disposed between the image sensor unit 28 and the telecentric imaging unit 26 and incorporates an aperture stop (hereinafter referred to as a variable aperture) (reference number 44 in FIGS. 2 and 3) having a variable opening. . The variable aperture unit 27 is used for adjusting the amount of light input to the linear image sensor and adjusting the depth of field.

  The scanning unit 24 described above moves three-dimensionally in the directions of the X, Y, and Z axes as the carriage 20 moves, and the target plane area 48 also moves along with it. The movable range of the target plane area 48 includes the object space 15 and installation locations of various signs 35, 37, and 39 described later.

  An illumination device 32 is also fixed to the carriage 20. The illumination device 32 is for illuminating the target plane region 48 and has at least two separated light source units 34A and 34B. That is, the support frame 30 is fixed to the front end portion of the pedestal 22 on the carriage 20 in parallel with the XY plane, and the X-axis direction ends of the support frame 30 (that is, the optical axis 43 of the scanning unit 24). Two light source units 34A, 34B are mounted on the left and right sides. Each of these two light source units 34A and 34B has a columnar shape, that is, a linear (linear) shape.

  The linear light source units 34 </ b> A and 34 </ b> B are arranged in parallel with the Y axis, and emit illumination light toward the target plane region 48. One light source unit 34A is arranged in the vicinity of the optical axis 43 of the scanning unit 24, while the other light source unit 34B is arranged at a certain distance from the optical axis 43 in the X-axis direction. For this reason, one light source unit 34A illuminates the target plane area 48 from substantially the front, while the other light source unit 34B illuminates the target plane area 48 from an oblique direction along the horizontal plane. Since these light source units 34A and 34B (illumination device 32) move together with the scanning unit 24 by the movement of the carriage 20, the positional relationship is always kept constant with respect to the scanning unit 24. Therefore, even if the target plane region 48 moves, the target plane region 48 is always irradiated with a certain amount of illumination light from a certain direction.

  At a place different from the platform 12, a control device 38 having a computer, a power supply circuit, and other electric / electronic circuits is installed. The control device 38 is electrically connected to the Y track 18 and the electric self-propelled device in the carriage 20, the linear image sensor in the scanning unit 24, the illumination device 32, and the like through a signal cable and a power cable. . The control device 38 controls and controls the three-dimensional movement and position of the carriage 20 by driving and controlling the Y track 18 and the electric self-propelled device in the carriage 20 (and thus controlling the scanning of the subject by the scanning unit 24). Function). The control device 38 further inputs image data sequentially output from the linear image sensor in the scanning unit 24 while the subject (not shown) is being scanned by the scanning unit 24, and processes the image data. By doing so, it has a function of synthesizing the focused image data of the subject. Further, the control device 38 has a function of performing shading correction associated with the above-described image processing, individual variable control of the output light amount of the light source units 34A and 34B, and various other incidental processes.

  A tilt check mark 35, a focusing mark 37, a white reference mark 39, and the like are located at positions far from the subject space 15 at the end of the table 14 on the far side (that is, the side facing the scanning unit 24) in FIG. Different types of signs are attached. The tilt check mark 35 displays, for example, a straight line parallel to the Y axis, and detects the tilt when the linear image sensor in the scanner unit 24 is slightly tilted with respect to the Y axis, as will be described later. Then, it is used in a process of setting a correction value for correcting this (step 104 in FIG. 16). The focus indicator 37 displays, for example, a fine lattice pattern (ladder chart) that cannot be read by the image sensor unless the focus of the scanning unit 24 is in focus. The unit 24 is used for processing (FIG. 16, step 106) for determining the origin of position control of the carriage 20 by adjusting the focus of the unit 24 to the mark 37. The white reference mark 39 displays, for example, a white reference color having a predetermined brightness (light reflectance), and is used in a process of setting a white reference level for shading correction (step 108 in FIG. 16). Here, for the white reference sign 39, different white reference colors having different brightness (light reflectance) are prepared (a plurality of separate white reference signs 39 for displaying different white reference colors are prepared. Or different white reference colors may be displayed on the same white reference sign 39). When performing the white reference level setting process, the user arbitrarily selects one white reference color from among the plurality of white reference colors according to the circumstances such as the color of the subject and the brightness of the desired image. Can be used. For example, if the subject is an overall bright color, the white reference color of pure white or higher brightness is used, while if the subject is an overall dark color, the white reference color of lower brightness is used. By using, it becomes easy to obtain an image with appropriate brightness according to the brightness of the subject.

  FIG. 2 shows the internal configuration of the scanning unit 24. FIG. 3 shows the positional relationship among the internal components of the scanning unit 24, the light source units 34 </ b> A and 34 </ b> B, and the target plane area 48.

  As shown in FIG. 2, the telecentric lens system 40 in the telecentric imaging unit 26 includes an objective lens 41 and a second lens 42, and as shown in FIG. The principal ray of the light travels parallel to the optical axis 43. As shown in FIG. 3, the telecetic lens system 40 transfers an image of the target plane area 48 at a predetermined distance 54, for example, 300 mm, forward from the objective lens 41 onto the linear image sensor 46 in the sensor unit 28. Form an image. The target plane area 48 is a linear area parallel to the Y axis, and is orthogonal to the optical axis 43 of the telecetric lens system 40 at the center point thereof.

  In such a telecentric lens system 40, in general, the objective lens 41 needs to be a large lens whose diameter is at least equal to the length of the target plane region 48 in the Y-axis direction. For example, when the length of the target plane region 48 in the Y-axis direction is set to 150 mm, the objective lens 40 needs to have a large diameter of 150 mm or more. However, if such a large lens is used as it is as the objective lens 41, the objective lens 41 may interfere with the attachment of the scanning unit 24 and the light source units 34A and 34B to the optimum position of the carriage 22. Therefore, in this embodiment, as shown in FIGS. 2 and 3, only a strip-shaped portion having a size that does not cause “cracking” is cut out from the circular lens having a large diameter as described above. The strip portion is arranged parallel to the Y axis and used as the objective lens 41. Since the strip-shaped objective lens 41 is particularly compact in the X-axis direction, there is no problem of obstructing the mounting of the scanning unit 24 and the light source units 34A and 34B as described above. In addition, by making the objective lens 41 of the telecentric lens system 40 into a strip shape, a reduction in lens manufacturing cost and a reduction in the weight of the scanning unit 24 are realized. The weight reduction of the scanning unit 24 also leads to the weight reduction of the X track 16, the Y track 18, and the carriage 20, and as a result, the weight of the entire apparatus is realized.

  A variable diaphragm 44 is disposed between the telesetric lens system 40 and the linear image sensor 46. The opening of the variable aperture 44 ranges from a completely closed state (blocking all incident light from the telecetic lens system 40) to a fully opened state (passing all incident light from the telecetic lens system 40). It is possible to variably set a plurality of levels, for example, 6 levels. The opening degree of the variable throttle 44 may be set by a user directly operating the variable throttle 44 manually or automatically by a command from the control device 38 shown in FIG. You may come to be.

The linear image sensor 46 converts, for example, a light conversion element array of, for example, 3000 pixels or more arranged in a straight line in parallel with the Y axis at a pitch of 8 μm, and an analog voltage signal output from the light conversion element array into digital pixel data. A / D conversion device. The A / D conversion device has an A / D conversion accuracy of, for example, 10 bits (1024 gradations) or more, and can cope with shading correction described later.
The magnification of the telecentric lens system 40 is, for example, 0.16. Accordingly, one section 49 of, for example, 50 μm (20 pixels / mm) on the target plane region 48 corresponds to one pixel of, for example, 8 μm on the linear image sensor 46. Therefore, for example, an image of the target plane region 48 having a length 58 of 150 mm is simultaneously formed on the linear image sensor 46 having 3000 pixels or more (50 μm × 3000 pixels = 150 mm). Since the effective length 58 of the target plane area 48 that can be simultaneously imaged is 150 mm, for example, as shown in FIG. 3, the lengths of the focusing mark 37 and the white reference mark 39 are also the minimum and the effective length 58, for example, 150 mm, Set to The length of the inclination check mark 35 is set to be longer than the effective length 58 of the target plane region 48, for example, twice or more, 300 mm, as will be described later.

  By using the telecentric lens system 40 as an imaging optical system, it is possible to obtain an orthogonal projection image of the subject. When a normal lens system that is not a telecentric system is used, the captured image is a central projection image in which the one close to the imaging system is large and the one that is far is small, and an orthographic projection image cannot be obtained. . By obtaining an orthographic projection image of the subject, it is possible to synthesize a focused image that faithfully represents the size and shape of the subject.

  The movable distance (stroke) of the scanning unit 24 is, for example, 500 mm in the X-axis direction, 1000 mm in the Y-axis direction, and 250 mm in the Z-axis direction. As a result, the maximum size of a subject whose whole image can be scanned is about 500 mm in diameter and about 1000 mm in height in the case of a three-dimensional subject having a substantially circular horizontal cross section, such as a basket, flower vase, or earthenware.

  As shown in FIG. 3, one linear light source unit 34A is arranged on the right side (positive direction of the X axis) of the objective lens 41 toward the target plane region 48, and the other linear light source unit 34A is arranged on the left side (negative direction of the X axis). A light source unit 34B is arranged. The right linear light source unit 34A irradiates the target plane area 48 (that is, the subject) with illumination light from substantially the front. This illumination light produces an effect that does not make the unevenness of the surface of the subject conspicuous. On the other hand, the left linear light source unit 34B irradiates the target plane region 48 (that is, the subject) with illumination light from the side surface direction. This illumination light has an effect of conspicuous the unevenness of the subject. A desired effect can be obtained by using or combining these two types of illumination light according to the subject. As an example, suppose you are shooting archaeological stoneware or earthenware. For example, in the case of an object that is black as a whole, such as obsidian stoneware, and the image does not easily reach contrast even if there is unevenness, the contrast according to the unevenness can be clarified by using illumination light from the side. . In addition, in the case of a subject that has a concave part toward the optical axis 43, such as Jomon pottery, use the illumination light from the front to illuminate the concave part deeply and photograph clearly. Can do. Further, in the case of a subject whose both side portions are located farther from the center, such as a cylinder, both the right and left linear light source units 34A and 34B are used in combination in order to reliably capture the contour. In such a subject, it is difficult for the illumination light to reach the opposite side only with illumination light from one side, and the image information of the entire accurate contour line cannot be obtained. The light quantity of each light source unit 34A, 34B can be adjusted separately so that a desired effect may be acquired. As described above, even if the illumination light quantity by the light source units 34A and 34B is changed, by performing shading correction described later, as long as the output of the linear image sensor 46 is not saturated, there is a gradation characteristic faithful to the reflectance of the subject. Images can be obtained many times with good reproducibility.

  The linear light source units 34A and 34B are both arranged parallel to the Y axis. A plane including two line-of-sight lines 50A and 52A connecting both ends of the linear light source unit 34A on the right side in the Y-axis direction and both ends of the target plane region 48 in the Y-axis direction (this plane is vertical; ) When assuming 53A, the illumination light emitted from the linear light source unit 34A travels from the linear light source unit 34A so as to enter the target plane region 48 while traveling in a narrow constant angle range substantially along the vertical surface 53A. The emission direction of the illumination light along the horizontal plane (XZ plane) is restricted to the narrow fixed angle range. In addition, the linear light source is set such that the distribution range of the incident angle along the vertical plane 53A of the illumination light incident on the target plane area 48 from the linear light source unit 34A is a narrow constant angle range at any location of the target plane area 48. The range of the emission angle of the illumination light from the unit 34A along the vertical plane 53A is restricted to the narrow constant angle range.

  Similarly, a plane 53B (hereinafter referred to as a vertical plane) including two line-of-sight lines 50B and 52B connecting both ends of the left linear light source unit 34B in the Y-axis direction and both ends of the target plane region 48 in the Y-axis direction. Assuming that the illumination light emitted from the linear light source unit 34B travels in a narrow fixed angle range along the vertical plane 53B and enters the target plane region 48, the illumination light from the linear light source unit 34B The emission direction along the horizontal plane (X-Z plane) is regulated within the narrow fixed angle range. In addition, the linear light source is set so that the distribution range of the incident angle along the vertical plane 53B of the illumination light incident on the target plane area 48 from the linear light source unit 34B becomes a narrow constant angle range in every part of the target plane area 48. The range of the emission angle along the vertical plane 53A of the illumination light from the unit 34B is restricted to the narrow constant angle range.

  The regulation of the emission direction of illumination light from the linear light source units 34A and 34B will be described more specifically with reference to FIGS.

  FIG. 4 is a cross-sectional view of the linear light source units 34A and 34B and the target plane region 48 cut along a horizontal plane.

  As shown in FIG. 4, the linear light source units 34A and 34B include a columnar or linear light source device 60A and 60B, respectively, and a columnar or linear elliptical reflecting mirror 62A disposed in parallel to the light source devices 60A and 60B, respectively. 62B. One focal axis of each of the elliptical reflecting mirrors 62A and 62B is disposed at the position of the central axis of each of the light source devices 60A and 60B, and the other focal axis is disposed at the position of the central axis of the target plane region 48. Accordingly, the distribution range of the incident angle of the illumination light incident on the target plane region 48 from the light source devices 60A and 60B is constant along the vertical planes 53A and 53B described above at every point in the Y direction of the target plane region 48. It is limited to a narrow angle range θA, θB.

  FIG. 5 is a cross-sectional view of one linear light source unit 34A or 34B and the target plane region 48 cut along the above-described vertical plane 53A or 53B (for convenience of explanation, the left side and right side in the reference numbers are distinguished. Index A and B are omitted).

  As shown in FIG. 5, the distribution range of the emission angle along the vertical plane 53 of the illumination light from the linear light source unit 34 is any point in the Y-axis direction of the portion (effective portion) that emits the light from the light source unit 34. In FIG. 2, the angle is restricted to a narrow fixed angle range (horizontal vicinity range) φ centered on the horizontal. There are several methods for regulating the emission angle, as will be introduced later.

Further, the length (effective length) of the effective portion of the light source unit 34 in the Y-axis direction is set as follows. That is, if this effective length is L1, the length of the target plane region 48 in the Y-axis direction is L3, and the distance between the target plane region 48 and the light source device 60 is D, the effective length L1 of the light source unit 34 is L1 ≧ L3 + L2
Here, L2 = 2 · D · tan (φ / 2),
Is set to satisfy the condition. For example, when L3 = 150 mm, D = 300 mm, and φ = 10 degrees, L2 = about 52 mm, and therefore, L1 ≧ 202 mm is set. In other words, the angle formed by the two line-of-sight lines 50 and 52 connecting both ends of the effective portion of the light source unit 34 and both ends of the target plane region 48 is equal to or greater than the angle φ (if When the line-of-sight lines 50 and 52 have an angle less than the angle φ, the illumination light incident angle range is less than the angle range φ, particularly near the both ends of the target plane region 48, and the portions close to the both ends. The amount of incident light is less than other parts).

  Under the configuration as described above, the distribution range of the incident angle along the vertical plane 53 of the illumination light incident on the target plane region 48 from the linear light source unit 34 is any point in the Y-axis direction of the target plane region 48. In this case, the light angle of the illumination light is constant at every point in the Y-axis direction of the target plane region 48. As a result, as will be described in detail later with reference to FIGS. 12 to 14, the lighting condition for all parts of the subject becomes uniform, and the entire image of the subject can be obtained with high image quality.

  Incidentally, when the emission angle of the illumination light from the light source unit 34 is not restricted as described above, the range of the incident angle of the illumination light to the target plane region 48 is as shown by the dotted line in FIG. From each point on the plane area 48, a wide angle range is set in consideration of the total length of the effective portion of the light source unit 34. Therefore, the incident angle range varies depending on the location on the target plane region 48, and the amount of incident light varies depending on the location. In this case, as will be described in detail later with reference to FIGS. 12 to 14, when attempting to obtain the entire image of the subject, there is an adverse effect of shadows on the uneven portions of the subject, which is high. It becomes difficult to obtain an image with an image quality.

  FIG. 6 is a cross-sectional view illustrating a specific configuration example of the light source device 60 in which the emission direction of the illumination light is limited within a certain angle range φ as described above.

As shown in FIG. 6, a light control device 62 is provided in front of the entire area of the light emitting device 60 that emits light from a columnar or linear light source device 60. The light control device 62 is, for example, a sheet or a plate made of a synthetic resin or other material having a high light transmission property, and includes a large number of light absorbing plates 64 (for example, a black matte plate) that absorbs light. Just like louver shading plates, they are arranged in parallel at regular intervals (for example, millimeter-order intervals). The sheet-like or plate-like light control device 62 is stretched in front of the light emitting surface of the light source device 60 in parallel with the light emitting device 60 or attached to the light emitting surface. The light source device 60 can be a linear lamp (for example, a rod-shaped fluorescent lamp). Of the illumination light emitted from the light exit surface of the light source device 60, the illumination light whose exit angle (angle with respect to the normal line of the light exit surface) is larger than a certain level hits the reflective layer 64 in the light control device 62. Absorbed. As a result, as shown in FIG. 6, the illumination light that can pass through the light control device 62 is limited in its emission angle within a certain narrow angular range φ centered on the horizontal. Here, the angle range φ depends on the thickness W of the light control device 62 and the distance d between the reflective layers 64.
φ = 2 · tan-1 (d / W)
Determined by

  7 and 8 are cross-sectional views illustrating another specific configuration example of the light source device 60 in which the emission direction of the illumination light is limited within a certain angular range φ. FIG. 7 is an exploded perspective view of the light guide used in this configuration example. FIG. 8 shows the overall structure and function of this configuration.

  In this configuration example, a columnar or linear light guide 65 as shown in FIGS. 7 and 8 is used. The light guide 65 includes a light guide rod 66 for guiding light to a columnar or linear region, and a light guide. The prism plate 68 is used to convert the direction of light emitted from the optical rod 66 into a predetermined direction. The light guide rod 66 is made of a synthetic resin or other material having a high light transmittance (for example, made of acrylic, polycarbonate, or optical glass), and as shown in FIGS. It has a quadrangular prism shape in which the angle formed by 66B and the rear surface 66C is an acute angle. The prism plate 68 is made of a synthetic resin or other material having a high light transmittance similar to that of the light guide rod 66 (for example, made of acrylic, polycarbonate, optical glass, etc.), and as shown in FIGS. The front surface 68A is flat, and the rear surface has a saw-tooth or washing board shape in which a large number of fine asymmetric triangular prism-shaped projections (prisms) 68B are continuously arranged in a line.

  As shown in FIG. 8A, the light guide rod 66 and the prism plate 68 are combined such that the front surface 66B of the former and the rear surface of the latter (the surface on which the projections and depressions of the columnar prism 68B are projected) are combined, thereby providing a linear guide. A light body 65 is formed. A lamp unit 64 including a halogen lamp and a concave reflecting mirror is attached to the base end portion (the end portion on the top surface 66A side of the light guide rod 66) of the light guide body 65. The light output from the lamp unit 64 enters the light guide rod 66 through the top surface 66 </ b> A of the light guide rod 66. The top surface 66A is processed to have fine irregularities, and the fine irregularities diffuse light passing therethrough in various directions. The light beam 70 that has entered the light guide rod 66 through the top surface 66A proceeds in the longitudinal direction of the light guide rod 66 while being repeatedly totally reflected by the front surface 66B and the rear surface 66C of the light guide rod 66, and the front surface 66B. The incident angle to the rear surface 66C gradually decreases, and finally, when the incident angle to the front surface 66B becomes smaller than the critical angle, the light passes through the front surface 66B and is emitted toward the prism plate 68. . Since the direction of light entering the light guide rod 66 from the top surface 66A varies due to the diffusion action of the top surface 66A, the amount of light emitted from the front surface of the light guide rod 66 is the length of the light guide rod 66. It is almost uniform over the entire range.

As shown in FIG. 8B, the light beam 70 emitted from the front surface of the light guide rod 66 enters the prism 68B from the upper surface of the prism 68B, is totally reflected by the lower surface of the prism 68B, and becomes a substantially horizontal light beam. It is ejected from the front surface 68A of the plate 68. The distribution of the emission angles of light emitted from the front surface 68A of the prism plate 68 is within a certain narrow angle range φ centered on the horizontal.
9 actually creates the light source device 60 having the configuration shown in FIG. 8, and sets the range of the emission angle along the vertical plane immediately before the light emission surface (the front surface 68A of the prism plate 68 in FIG. 8). The measurement results are shown. Here, regarding the emission angle, the direction of the normal line of the light emission surface is set to 0 degree.

  As can be seen from the graph 72 showing the emission angle distribution in FIG. 9, the emission angle is limited to a very narrow angle range. For example, if the range in which the radiation intensity is half of the maximum value is defined as the aforementioned angle range φ, φ is about 10 degrees.

  FIG. 10 is a perspective view for explaining a method for three-dimensionally scanning the three-dimensional subject arranged in the subject space 15 shown in FIG. 1 by the scanning unit 24.

  As shown in FIG. 10, the space where the three-dimensional subject 80 exists is divided into a plurality of (N) layers having different Z coordinates. The thickness of each layer 82-1, 82-2,..., 82-N is equal to or less than the depth of field of the scanning unit 24. As the scanning unit 24 moves in the Z-axis direction at a pitch corresponding to the thickness of the layer, the target plane area 48 in front of the scanning unit 24 becomes the Z of each of the layers 82-1, 82-2, ..., 82-N. Sequentially arranged in coordinates. The scanning unit 24 moves in the X- and Y-axis directions at the Z coordinate of each layer 82-i (i = 1, 2,..., N), so that the target plane region 48 moves through each layer 82-i. In the meantime, image data sequentially output from the scanning unit 24 is collected by the control device 38, and as a result, image data of each layer 82-i is obtained in the control device 38. It is done. The image data of each layer 82-i includes in-focus pixel data focused on the subject 80 at a position where each layer 82-i and the subject 80 intersect. By performing the above operation for all the layers 82-1, 82-2,..., 82-N, the image data of all the layers 82-1, 82-2,. Is obtained. In the control device 38, focused pixel data focused on the subject 80 is extracted from the image data of all the layers 82-1, 82-2,..., 82-N, and the focused pixel data is collected. Thus, in-focus image data representing the entire image of the subject 80 in focus is synthesized at all locations on the subject 80.

  When the dimension (height) of the subject 80 in the Y-axis direction is larger than the length (effective length) (for example, 150 mm) of the target plane region 48 in the Y-axis direction, naturally the dimension of each layer 82-i in the Y-axis direction. Becomes larger than the effective length of the target plane region 48. In this case, each layer 82-i is divided into a plurality of bands 84-1, 84-2,..., 84-M each having a width in the Y-axis direction corresponding to the effective length of the target plane region 48. The scanning operation of each layer 82-i by the target plane region 48 is decomposed into a scanning operation of a plurality of bands 84-1, 84-2,..., 84-M constituting each layer 82-i. That is, the scanning unit 24 moves in the Y-axis direction at a pitch corresponding to the effective length of the target plane region 48, so that the target plane region 48 is in each of the bands 84-1, 84-2, ..., 84-M. Sequentially placed on the Y coordinate. Then, the scanning unit 24 moves in the X-axis direction at the Y coordinate of each band 84-j (j = 1, 2,..., M), so that the target plane region 48 moves each band 84-j over the entire area. In the meantime, image data sequentially output from the scanning unit 24 is collected by the control device 38. As a result, image data of each band 84-j is obtained in the control device 38. By performing the above-described operation for all the bands 84-1, 84-2,..., 84-M of each layer 82-i, all the bands 84-1, 84-2,. , 84-M image data is obtained in the control device 38. Within the control device 38, the image data of all the layers 82-1, 84-2,..., 84-M of each layer 82-i are joined in the Y-axis direction to obtain the image data of each layer 82-i. It is done.

  FIG. 11 is a diagram for explaining a processing method for extracting focused pixel data focused on the subject 80 from the image data of the plurality of layers 82-1, 82-2,..., 82-N shown in FIG. It is.

FIG. 11A is a partial cross-sectional view of the subject 80 cut along an XZ plane (horizontal plane) passing through the Y coordinate j. The hatched portion represents the inside of the subject 80, and the line 86 represents the contour of the subject 80. Here, as an example, seven pixels located at the same X coordinate i + 4 and having different layers, that is,
Pixel P (i + 4, j, k) of layer (k),
Pixel P (i + 4, j, k + 1) of layer (k + 1),
Pixel P (i + 4, j, k + 2) of layer (k + 2),
Pixel P (i + 4, j, k + 3) of layer (k + 3),
Pixel P (i + 4, j, k + 4) of layer (k + 4),
Pixel P (i + 4, j, k + 5) of layer (k + 5),
Pixel P (i + 4, j, k + 6) of layer (k + 6),
Pay attention to. The graph 88 of FIG. 11B shows the result of calculating a predetermined focusing function for these seven pixels. Here, the in-focus function is a function whose value changes depending on the degree of focus. For example, the contrast value at the target pixel P (a, b, c) in the image of the layer is adopted. Can do. As a contrast value, for example,
Contrast (a, b, c) = max 3 (a, b, c) −min 3 (a, b, c)
Can be used. Here, max3 (a, b, c) and min3 (a, b, c) are 9 pixels in a 3 × 3 pixel region centered on the pixel of interest P (a, b, c) in the layer image. Are the functions for obtaining the maximum value and the minimum value from the tone values. As shown in the graph 88 of FIG. 11B, among the pixels of different layers (Z coordinates) having the same X and Y coordinates, the pixel P (i + 4, j, k + 4) having the maximum focusing function is the subject. Extracted as the in-focus pixel with the best focus.

  As described above, a process of calculating and comparing in-focus functions of pixels in different layers (Z coordinates) having the same X and Y coordinates, and extracting one pixel having the maximum in-focus function as the in-focus pixel. Is repeated for all X, Y coordinates in the layer. The extracted in-focus pixels are collected for all the X and Y coordinates, and one image is synthesized. The synthesized image is an orthographic projection image obtained by orthogonally projecting a subject having a three-dimensional structure on the XY plane, and is a focused image in which the entire area of the projected subject is in focus.

  12 to 15 illustrate the advantages obtained by the illumination light direction restriction described with reference to FIGS. 5 to 9 under the scanning method described with reference to FIG. 10.

  FIG. 12 shows a typical example of a subject shape in which the advantage of restricting the direction of illumination light appears most.

  As shown in FIG. 12A, an object 90 having a protrusion 92 extending obliquely with respect to the Y axis and the X axis on the surface thereof is assumed here. Then, as shown in FIG. 12B, when scanning the subject 90, a case is assumed in which the boundary between two adjacent bands B1 and B2 crosses the protrusion 92 obliquely. In this case, attention is paid to the four points P1, P2, P3, and P4 located immediately above and below the protrusion 92 on the surface of the subject 90. Among these, the points P1 and P2 are photographed when the lower band B1 is scanned, and the points P3 and P4 are photographed when the upper band B2 is scanned.

  FIG. 13 shows the illumination light to the points P1, P2, P3, and P4 in the case of FIG. 12 described above when no restriction is imposed on the emission direction of the illumination light from the linear light source unit 34. The distribution of the incident angles is shown.

  In this case, the illumination light is incident on every point of the subject 90 in an angle range in which the entire length of the light source unit 34 is viewed from each point. Therefore, at the time of scanning the lower band B1, as shown in FIG. 13A, the point P1 directly below the ridge 92 is not compared with the length of the light source unit 34 that does not fall behind the ridge 92 as viewed from the point P1. The illumination light is incident from the long part, mainly illuminating from below. Therefore, the incident angle range λ1 to the point P1 is relatively wide, and the amount of incident light is relatively large. On the other hand, the point P2 directly above the ridge 92 is illuminated so as to illuminate mainly from the upper side from a relatively short portion of the total length of the light source unit 34 that does not enter the ridge 92 when viewed from the point P2. Light enters. Therefore, even if the brightness of the points P1 and P2 of the subject 90 is the same, in the obtained image of the band B1, the image of the point P2 immediately above the ridge 92 is darker than the point P1 directly below.

  Further, during scanning of the upper band B2, as shown in FIG. 13B, the incident angle range λ3 to the point P3 immediately below the protrusion 92 is relatively narrow and the amount of incident light is relatively small. The incident angle range λ4 to the point P4 immediately above is relatively wide, and the amount of incident light is relatively large. As a result, even if the brightness of the points P3 and P4 of the subject 90 is the same, in the obtained image of the band B2, the image of the point P3 immediately below the ridge 92 is darker than the point P4 directly above. Therefore, in the image of the band B2, the point P3 immediately below the ridge 92 is dark, but the point P4 immediately above is bright.

  As a result, when the images of the bands B1 and B2 are joined, as shown in FIG. 15A, a shadow 94 is formed on the upper side of the pier 92 in the region of the band B1, and a shadow is formed on the lower side of the pier 92 in the region of the band B2. 96, and a low-quality image in which the brightness is mismatched at the joint between the bands B1 and B2 can be obtained.

  FIG. 14 shows the illumination to the points P1, P2, P3, and P4 when the emission direction of the illumination light from the linear light source unit 34 is restricted as described in FIG. 5 in the case of FIG. The distribution of the incident angle of light is shown.

  As shown in FIG. 14A and FIG. 14B, when scanning any of the bands B1 and B2, any point such as the points P1, P2, P3, and P4 is constant from the light source unit 34 in the horizontal direction. Illumination light is incident in a narrow incident angle range φ, and therefore the amount of incident light is constant at any point. As a result, as shown in FIG. 15B, a high-quality image faithful to the brightness of the subject 90 without the shadows 94 and 96 as shown in FIG. 15A can be obtained.

  FIG. 16 shows the overall flow of the scanning operation control performed by the control device 38 shown in FIG.

  As shown in FIG. 16, in step 100, the origin of the moving mechanism (X trajectory 16, Y trajectory 18, and carriage 120) is detected. That is, the scanning unit 24 is moved to the mechanical origin of the X, Y, and Z axes of the moving mechanism. In step 102, it is checked whether the scanning operation to be performed is the first scanning operation after the three-dimensional object image scanner 10 is assembled or the second and subsequent scanning operations. As a result of this check, if it is determined that the second time or later, the control proceeds to step 108, but if it is determined to be the first time, steps 104 and 106 are performed before proceeding to step 108.

  In step 104, a correction value is set to correct the inclination of the line sensor (rotation angle about the optical axis 43 with respect to the Y axis).

  FIG. 17 shows a specific flow of the tilt correction value setting process in step 104. FIG. 18 explains the meaning of this processing in an easy-to-understand manner.

  As shown in FIG. 17, in step 120, the target area 48 is aligned with the inclination check mark 35 shown in FIG. 1 and the line chart (parallel to the Y axis) displayed on the inclination check sign 35 is displayed. Image) is read by two band scans. That is, as shown in FIG. 18A, the lower half of the line chart 130 on the inclination check mark 35 is read by the lower band scan 132A, and the upper half is read by the upper half band scan 132B. At this time, if the linear image sensor 46 is tilted with respect to the Y-axis, the target plane region 48 is similarly tilted as shown in FIG. As shown in FIG. 18B, the image of the line chart 130 becomes two inclined discontinuous lines 134A and 134B.

  In step 122 of FIG. 17, particularly the band joint portion of the image of the line chart 130 as shown in FIG. 18B is enlarged and displayed on the monitor (not shown) of the control device 38. At this time, not only the image showing the reading result shown in FIG. 18B as it is, but also different correction values (shear deformation amounts) for the upper line 134A and the lower line 134B as shown in FIGS. 18C and 18D. ) A plurality of corrected images subjected to shear deformation in S1 and S2 are also displayed together with their band joint portions enlarged.

  In step 124 of FIG. 17, the user selects an optimal line chart image in which the upper and lower lines 134A and 134B are connected in a straight line from among a plurality of images displayed on the monitor. For example, among the images shown in FIGS. 18B, 18C, and 18D, the image of FIG. 18D is selected as the optimum one.

  In step 126 in FIG. 17, the correction value (shear deformation amount with respect to the image from the line sensor) (for example, S2 shown in FIG. 18D) applied to the selected optimum image is sent to the control device 38 as the inclination correction value setting. Remembered.

  Reference is again made to the overall control flow of FIG. After the setting of the tilt correction value in step 104 described above is completed, lens focus adjustment is performed in step 106. That is, the telecentric lens system 40 and the linear image sensor 46 so that the ladder chart displayed on the focusing mark 37 shown in FIG. 1 can be clearly read with the Z coordinate of the moving mechanism positioned at the preset origin. The distance between them or the focal length of the telecentric lens system 40 is fine tuned so that the target area 48 is accurately aligned with the Z coordinate of the focus indicator 37. Alternatively, in place of the fine adjustment of the distance, focusing may be performed by a method of setting the Z coordinate of the moving mechanism when the ladder chart can be read clearly as the Z coordinate origin.

  In step 108, a scanning range, that is, a range of X, Y, and Z coordinates in which three-dimensional scanning is performed is set. For example, a range from the origin to X_set, Y_set, and Z_set is set. In step 110, a white reference level and a black reference level for shading correction are set.

  FIG. 19 shows the method of setting the white / black reference level in step 110 in an easy-to-understand manner. FIG. 20 shows a specific flow of this processing.

  As shown in FIG. 19, when the output voltage level when the linear image sensor 46 is saturated is Vsat and the output voltage level when there is no input (dark) is Vbk, Vsa / Vbk can be obtained from the linear image sensor 46. The S / N ratio is generally about 100 to 1000. The gradation which is one of the factors of the image quality becomes better as the S / N ratio is larger. When the linear image sensor 46 is saturated, the gradation reproducibility of the image is lost, so that an operation region lower than the saturation output level Vsat is used. Factors that change the output voltage level of the linear image sensor 46 include the light amount of the light source unit 34, the F number of the lens system 40 (the opening of the variable aperture 44), the sensitivity of the linear image sensor 46, and the charge accumulation time of the linear image sensor 46. , The brightness (reflectance) of the white reference sign.

  As shown in step 158 of FIG. 20, when setting the white reference level, the white reference mark 39 shown in FIG. 1 is read, and the output voltage level Vwh of the linear image sensor 46 at that time is measured for each pixel. Is done. At that time, the light amount of the light source unit 34 and the lens system 40 are adjusted so that the output voltage level Vwh of the linear image sensor 46 is close to and not exceeding the saturation output voltage level Vsat (FIG. 20, step 160). The F number (opening of the variable aperture 44) and the charge accumulation time of the linear image sensor 46 are adjusted manually or automatically (FIGS. 20, steps 152, 154, 156). For example, if the saturation output voltage level Vsat is 1024 as a digital output level, the above-described light amount, F number (throttle opening), and charge so that the output voltage level Vwh at the time of reading the white reference sign is about 1000. The accumulation time is adjusted. When the adjustment is completed as described above, the output voltage level Vwh of the linear image sensor 46 is measured for each pixel and stored in the control device 38 as the white reference level setting (step 162 in FIG. 20).

As already described, a plurality of white reference signs 39 having different brightness (reflectance) are prepared. When setting the white reference level, a white reference sign having an appropriate reflectance is selected from the plurality of white reference signs 39 according to the average brightness (reflectance) of the subject. Used (FIG. 20, step 150). For example, a white reference mark having a reflectance close to and exceeding the average brightness (reflectance) of the subject is selected. For example, if a subject has an overall color close to black, it is brighter to use a relatively dark white reference sign that has a reflectivity close to, but not less than, the subject's low reflectivity. Compared with the case where a white reference mark is used, an image with a high S / N ratio that is less affected by the dark noise of the linear image sensor 46 can be obtained.
When the setting of the white reference level is completed, the black reference level is set in steps 164 to 168 in FIG. In this process, after the light source unit 34 is turned off and the variable aperture 44 is completely closed (step 164), the dark part output voltage level Vbk of the linear image sensor 46 is measured for each pixel (step 166), and the measurement is performed. The dark part output voltage level Vbk for each pixel is stored in the control device 38 as a black reference level setting (step 168).

  After the black reference level is set, the light source unit 34 is turned on again in step 170 of FIG. 20, and the light quantity of the light source unit 34 and the opening of the variable aperture 44 are adjusted with the white reference level setting. Returned to the value at the time of manual or automatic.

  Reference is again made to the overall control flow of FIG. After the setting of the black / white reference level in step 110 is completed, the subject is three-dimensionally scanned (steps 112 to 116). When the three-dimensional scanning is completed, a focused image of the subject is synthesized in step 118 based on the image data acquired by the three-dimensional scanning.

  FIG. 21 shows a specific control flow of three-dimensional scanning and in-focus image synthesis in steps 112 to 118 described above.

  As shown in FIG. 21, in step 182, the movement pitch (Z pitch) in the Z-axis direction (that is, the layer 82 shown in FIG. 10) according to the F number of the telecentric lens system 70 (the opening degree of the variable aperture 44). -1, 82-2,..., 82-N) is set to be equal to or less than the depth of field. In addition, the F number (aperture opening) may be set so that the Z pitch is determined first and the depth of field becomes larger than that (in this case, in the white reference level setting process, the above-described setting is performed). Under the F number (aperture opening), the light source light amount and the charge accumulation time are adjusted so that an appropriate white reference level is obtained. Different Z pitches and F numbers (throttle opening) can be set according to the purpose of scanning and the state of the subject. For example, when aiming at high-speed scanning, or when scanning a subject whose shape is not completely fixed such as a plant, scan with a small number of layers under a large depth of field. The aperture can be narrowed to increase the F number, and the Z pitch can be set large. On the other hand, for the purpose of obtaining a high-definition subject image, for example, to increase the image illuminance, scan with fine Z pitch, widen the aperture to reduce the F number, and Z pitch also Can be set small.

In step 184, the carriage 20 is moved, and the X, Y, Z coordinates of the carriage 20 are arranged at a predetermined scanning start position (for example, the origin). In step 186, one line of image data where the target plane region 48 is located at the scanning start position is output from the scanning unit 24 and acquired by the control device 38. In step 190, shading correction for each pixel with respect to the acquired one-line image data is performed by the control device 38. The shading correction is performed using the white reference level and the black reference level set in step 110 in FIG.
d_data (m) = K ・ {D_data (m)-D_bk (m)} / {D_wh (m) -D_bk (m)}
Where d_data (m) is the data value after shading correction of pixel m
K is a constant,
D_data (m) is the original data value before shading correction of pixel m,
D_wh (m) is the white reference level,
D_bk (m) is the black reference level,
It is done by the method. By performing shading correction, the image data is always image data that is faithful to the reflectance of the subject without depending on illumination unevenness or illumination light quantity.

  In step 192, the carriage 20 moves in the X-axis direction by a predetermined X pitch (typically, one pixel size corresponding to the resolution of the desired image in the X-axis direction). Then, the operations in steps 186 and 188 are performed again at the position after the movement. Steps 186 to 192 described above are repeated until the X coordinate of the carriage 20 reaches the X coordinate X_set at the end of the scanning range set in step 108 of FIG. 16 (YES in step 190). When the terminal X coordinate X_set is reached (YES in step 190), the control proceeds to step 194, where the image data of a large number of lines acquired so far are collected and one band of image data is synthesized. Note that it is not always necessary to perform this process as a separate process in step 194, and the image data of the line obtained in steps 186 and 188 is recorded in the storage device in the control device 38 in association with the coordinate information. Naturally, band image data is generated on the storage device.

  Thereafter, in step 198, the carriage 20 is moved in the Y-axis direction by a predetermined Y pitch (typically, the effective length of the target plane area 48). Thus, the carriage 20 is arranged at the Y coordinate corresponding to the next band. In step 200, the carriage 20 is placed at the X coordinate corresponding to the scan start position of the band. In this step 200, if the X scan direction of the band is the same in all bands, the carriage 20 is returned to the first X coordinate (for example, the X coordinate origin), but on the other hand, If the direction is reverse between bands (that is, zigzag scanning), the carriage position when the carriage 20 moves to the Y coordinate of the next band is directly used as the scan start position of the next band.

  Thereafter, Steps 186 to 192 described above are repeated, and in Step 194, image data of the next band is obtained.

  Steps 186 to 200 described above are repeated until the Y coordinate of the carriage 20 reaches the Y coordinate Y_set at the end of the scanning range set in step 108 of FIG. 16 (YES in step 196). When the end Y coordinate Y_set is reached (YES in step 196), the control proceeds to step 202, where the image data of a plurality of bands acquired so far are joined in the Y-axis direction, and one layer of image data is obtained. Synthesized. At that time, the band image is subjected to the shear deformation correction using the setting S2 of the inclination correction value (shear deformation amount) as shown in FIG. 18D, and is thus influenced by the inclination of the linear image sensor 46. Thus, a layer image in which bands are connected in a consistent manner is obtained. This step 202 does not necessarily need to be performed as a separate process, and the band image data obtained in step 194 is associated with the coordinate information (coordinate information after shear deformation correction) and stored in the control device 38. By recording in the device, the image data of the layer is naturally generated on the storage device.

  Thereafter, in step 206, the carriage 20 moves in the Z-axis direction by a predetermined Z pitch (set in step 192). As a result, the carriage 20 is arranged at the Z coordinate corresponding to the next layer. In step 208, the carriage 20 is placed at the X and Y coordinates corresponding to the scan start position of the layer. In step 208, if the X and Y scan directions of the layers are the same in all layers, the carriage 20 is returned to the first X and Y coordinates (for example, the X and Y coordinate origins). On the other hand, when the direction is opposite between the adjacent layers, the position of the carriage 20 at the time when the carriage 20 moves to the Z coordinate of the next layer becomes the scan start position of the next layer as it is.

  Thereafter, the above-described steps 186 to 200 are repeated, and the image data of the next layer is obtained in step 202.

  Steps 186 to 208 described above are repeatedly performed until the Z coordinate of the carriage 20 reaches the Z coordinate Z_set at the end of the scanning range set in step 108 of FIG. 16 (YES in step 204). When the terminal Z coordinate Z_set is reached (YES in step 204), the control proceeds to step 210, where the subject is best focused on for each X and Y coordinate from the image data of a large number of layers acquired so far. In-focus pixel data is extracted. In the method for extracting focused pixel data for each X and Y coordinate, as already described with reference to FIG. 11, attention is paid to pixels having the same X and Y coordinates in a large number of layers, and a predetermined value is set for each pixel. A method of calculating the in-focus function (for example, contrast value), comparing the in-focus function values between those pixels, and extracting the data of the pixel with the largest in-focus function value as the in-focus pixel data is adopted. it can.

  In step 212, all the extracted in-focus pixel data of the X and Y coordinates are collected, and in-focus image data in which all the points of the subject are in focus are generated and output. At this time, preferably, the focused image data is subjected to gamma correction corresponding to the gamma characteristic of the linear image sensor 46 or an image output device (for example, a monitor, a printer) or the like, and thereby more faithful to the state of the subject. Correct image is output. This step 212 does not necessarily need to be performed as a separate process, and the in-focus pixel data (preferably gamma corrected) extracted in step 210 is associated with the coordinate information and stored in the control device 38. By being recorded in the apparatus, the focused image data is naturally generated on the storage device.

  As mentioned above, although embodiment of this invention was described, this embodiment is only the illustration for description of this invention, and is not the meaning which limits the scope of the present invention only to this embodiment. The present invention can be implemented in various other modes without departing from the gist thereof.

  For example, in the above-described embodiment, the positional relationship between the scanning unit 24 and the target plane area 48 is fixed, and when the scanning unit 24 is moved by the carriage 20, the target plane area 48 moves together and the subject is tertiary. Although originally designed to scan, this is just one example. As a modification, by changing the configuration of the optical system in the scanning unit 24, the positional relationship between the scanning unit 24 and the target plane region 48 changes, and the scanning unit 24 remains stationary or only moves in a small range. It is also possible for the target plane region 48 to move over a large range. In that case, in order to always give a certain amount of illumination light to the target plane region 48 from a certain direction, the illumination device 32 may be moved together with the target plane region 48.

  In the above-described embodiment, the three-dimensional scanning is performed in the order of movement in the X-axis direction, movement in the Y-axis direction, and movement in the Z-axis direction. However, the X-, Y-, and Z-axis directions are not necessarily in that order. There is no need to perform the movement, and other orders may be used. In the above embodiment, the Y axis is the vertical direction, but the Z axis may be the vertical direction, or one of the coordinate axes may be oblique to the vertical or horizontal. A two-axis scanner having only the X-axis and the Y-axis may be used.

  By the way, in the above-described embodiment, in the case of a subject whose both side portions are located farther from the center, such as a cylinder, the light source of the linear light source units 34A and 34B is irradiated on the right and left contours. The contribution rate is different. Here, when the illumination unevenness patterns of the linear light source unit 34A and the linear light source unit 34B are extremely different, in the above-described embodiment, even if shading correction is performed, the illumination unevenness (especially, the left and right linear light source units 34A, 34A, There is a case where the unevenness of illumination at the right end and the left end of the subject where the difference in the contribution ratio of the irradiation light of 34B becomes more remarkable cannot be satisfactorily corrected.

  FIG. 22 shows an overall flow of an example of scanning operation control performed by the control device 38 shown in FIG. 1 in order to improve this problem.

  The flow of Step 100, Step 102, Step 104, Step 106, and Step 108 is exactly the same as the scanning operation control shown in FIG. In step 310A, the white reference level and the black reference are set for shading correction. Here, only one linear light source unit (hereinafter referred to as the first linear light source unit) 34A is turned on, and the other one is turned on. The linear light source unit (hereinafter referred to as the second linear light source unit) 34B is not lit. The white reference level and black reference setting method other than the lamp control in which only one of the lamps is turned on in this manner is the same as the control in FIG. 16 described above. After the setting of the white / black reference level in step 310A is completed, the subject is three-dimensionally scanned (steps 312A to 316A). Again, only the first linear light source unit 34A that is lit when setting the white / black reference level is lit, and the second linear light source unit 34B is not lit. When the three-dimensional scanning is completed, based on the image data acquired by the three-dimensional scanning, in step 318A, the first sub-focus image of the subject (obtained only under the illumination of the first linear light source unit 34A). The in-focus image) is synthesized. The image data processing method for synthesizing the first sub-focus image is the same as that in the control of FIG. 16 described above.

  Next, in step 310B, only the second linear light source unit 34B that has not been lit previously is lit, and the first linear light source unit 34A that has been lit previously is not lit, and the white reference level for shading correction and The black reference is set. The white reference level and black reference setting method other than the lamp control for lighting only the second near light source unit 34B is the same as the control in FIG. After the setting of the white / black reference level in step 310B is completed, the subject is three-dimensionally scanned (steps 312B to 316B). Again, only the second linear light source unit 34B is lit, and the first linear light source unit 34A is not lit. When the three-dimensional scanning is completed, based on the image data obtained by the three-dimensional scanning, in step 318B, the second sub-focus image of the subject (obtained only under illumination of the second linear light source unit 34B). The in-focus image) is synthesized. The image data processing method for synthesizing the second sub-focus image is the same as that in the control of FIG. 16 described above. Thereafter, in step 320, the pixels at the same image position corresponding to the first sub-focus image and the second sub-focus image are integrated (typically, averaging), and one final focus image is obtained. The focus image is synthesized and output.

  According to this control, even when the illumination unevenness patterns of the first and second linear light source units 34A and 34B are extremely different, shading correction is performed under illumination of only the respective lamps, and the sub-focus image is obtained. By creating and integrating both sub-focus images at the end, it is possible to correct illumination unevenness for any uneven subject, resulting in image data faithful to the reflectance of the subject .

  Note that FIG. 22 illustrates an example of control when two light sources are used. However, even when more light sources are used, each subject is illuminated under the illumination of each light source according to the control principle of FIG. Can be obtained, and finally, the images under different illuminations of the light sources can be integrated (for example, addition average), and one final image can be synthesized.

  In the above-described embodiment, the light source as the illumination means has moved in the three-dimensional direction together with the image sensor, but may be fixedly arranged illumination that does not move together with the image sensor. It may be an external light source that is separate from the image scanner according to the invention. For example, it is also possible to turn on only the indoor lighting in which the image scanner is installed and to turn off all the light sources mounted on the image scanner for imaging. By doing so, the degree of freedom of the illumination method for the subject is increased, and there is an advantage that the appearance of shadows and the like can be freely controlled. Therefore, if the external light source is set so that the illuminance distribution of the subject is in a desired state, a well-illuminated subject image can be obtained even if the control shown in FIG. 16 is performed without the control shown in FIG. be able to.

  As mentioned above, although embodiment of this invention was described, this embodiment is only the illustration for description of this invention, and is not the meaning which limits the scope of the present invention only to this embodiment. The present invention can be implemented in various other modes without departing from the gist thereof.

1 is a perspective view showing an overall configuration of a three-dimensional object image scanner according to an embodiment of the present invention. FIG. 4 is a cross-sectional view showing an internal configuration of a scanning unit 24. The perspective view which shows the positional relationship of the internal components of the scanning unit 24, light source unit 34A, 34B, and the target plane area | region 48. FIG. Sectional drawing which cut | disconnected linear light source unit 34A, 34B and the target plane area | region 4 along the horizontal surface. FIG. 6 is a cross-sectional view of the linear light source unit and the target plane region 48 cut along a plane 53. Sectional drawing which shows the specific structural example of the light source device 60 by which the emission direction of illumination light was restrict | limited within the fixed angle range (phi). Sectional drawing which shows the component of the linear light guide used by another specific structural example of the light source device 60 by which the emission direction of illumination light was restrict | limited within the fixed angle range (phi). Sectional drawing which shows another specific structural example of the light source device 60 using the light guide of FIG. The light source device 60 having the configuration shown in FIG. 8 was actually created, and the range of the emission angle along the vertical plane was measured immediately before the light emission plane (front surface 68A of the prism plate 68 in FIG. 8). FIG. It is a perspective view explaining the scanning method with respect to a three-dimensional photographic subject. FIG. 11 is a diagram for explaining a processing method for extracting focused pixel data focused on a subject 80 from the image data of a plurality of layers 82-1, 82-2,..., 82-N shown in FIG. FIG. 11 is a diagram illustrating a typical example of a subject shape in which the direction restriction of the illumination light illustrated in FIG. 5 appears most frequently under the scanning method illustrated in FIG. 10. The figure explaining the problem when the direction control of illumination light is not made. FIG. 14 is a diagram for explaining that the problem shown in FIG. 13 is solved by restricting the direction of illumination light. The figure which contrasts and shows the image obtained by image | photographing the to-be-photographed object of FIG. 10 with the case where the direction regulation of illumination light is not made | formed. 7 is a flowchart showing an overall flow of scanning operation control performed by the control device 38. 17 is a flowchart showing a specific flow of the tilt correction value setting process in step 104 of FIG. 16. The figure explaining the meaning of the process of FIG. FIG. 17 is a diagram for explaining the method of setting a white / black reference level in step 110 of FIG. 16 in an easy-to-understand manner. 5 is a flowchart showing a specific flow of processing for setting the white / black reference level. 17 is a flowchart showing a specific flow of three-dimensional scanning and in-focus image synthesis in steps 112 to 118 in FIG. 16. The flowchart which shows the whole flow of the scanning operation | movement control for eliminating the influence of the difference in the illumination nonuniformity pattern between right-and-left light source units performed by the control apparatus 38. FIG.

Explanation of symbols

14 Table 15 Object space 16 X orbit 18 Y orbit 20 Carriage 24 Scanning unit 26 Telecentric imaging unit 27 Aperture unit 28 Image sensor unit 32 Illumination device 34 Light source unit 35 Tilt check indicator 37 Focusing indicator 38 Control device 39 White reference indicator 40 Telecentric Lens System 43 Optical Axis 44 Variable Aperture 46 Linear Image Sensor 48 Target Plane Region 53 Vertical Plane 60 Light Source Device 62 Elliptic Reflector 63 Light Control Device 65 Light Guide 66 Light Guide Rod 68 Prism Plate 82 Layer 84 Band 80, 90 Subject 92 Projection

Claims (8)

In an image scanner for obtaining an orthographic projection image of a three-dimensional subject,
An image sensor;
A telecentric imaging system which is disposed between the image sensor and a target plane area remote from the image sensor and forms an image of the target plane area on the image sensor;
Moving means for moving the target plane area in a three-dimensional direction in a subject space in which a subject is arranged;
First and second light sources for irradiating illumination light to the target plane region, disposed at positions on both sides of an optical axis from the telecentric imaging system toward the target plane region;
Image processing means,
The contribution rates of the first and second light sources are different in the subject,
The image processing means includes
(A) Of the first and second light sources, only the first light source is turned on, and the target sensor is sequentially output from the image sensor while the target plane area is moving in the object space. The first image data corresponding to the reflectance by the first light source in the target plane area is input, and the pixel data in focus of the subject is detected from the input first image data. Collecting the detected pixel data to obtain and store a first sub-orthographic image,
(B) Only the second light source among the first and second light sources is turned on, and the target sensor is sequentially output from the image sensor while the target plane area is moving in the subject space. The second image data corresponding to the reflectance of the second light source in the target plane area is input, and the focused pixel data of the subject is detected from the input second image data. Collecting the detected pixel data to obtain and store a second sub-orthographic image,
(C) obtaining orthographic image according to the reflectivity of the object by integrating the first and second sub-orthographic image,
Solid object image scanner.   In (A), the image processing means detects a white reference level of the image sensor in a state where only the first light source among the first and second light sources is turned on, and detects the detected white Shading correction is performed on the first image data output from the image sensor using a reference level, and the input first image data is first image data after shading correction,
  In (B), the image processing means detects a white reference level of the image sensor in a state where only the second light source among the first and second light sources is lit, and detects the detected white Shading correction is performed on the second image data output from the image sensor using a reference level, and the input second image data is second image data after shading correction.
The three-dimensional object image scanner according to claim 1.   A plurality of white reference signs of different brightness that can be selectively used when detecting the white reference level of the image sensor;
  In (A) and (B), the image processing means selects a white reference mark corresponding to an average reflectance of the subject from the plurality of white reference signs, and based on the selected white reference sign Detecting a white reference level of the image sensor;
The three-dimensional object image scanner according to claim 2.   A scanning unit having the image sensor and the telecentric imaging system;
  The moving means moves the target plane region in a three-dimensional direction by moving the scanning unit,
  The target plane area is a linear plane area parallel to the Y direction perpendicular to the observation direction of the subject,
  The image sensor is a linear sensor arranged substantially parallel to the Y direction,
  The objective lens of the telecentric imaging system is a strip-like convex lens arranged in parallel with the target plane region,
  The strip-shaped convex lens has a shape obtained by cutting a circular objective lens having a diameter equal to or longer than the length in the Y direction of the target plane region into a strip shape, and the strip-shaped convex lens in the Y direction. The length along is not less than the length along the Y direction of the target plane region.
The three-dimensional object image scanner according to any one of claims 1 to 3. The target plane area is a linear area;
In the illumination means, each of the two light sources is a linear light source,
The linear light source is configured to move together with the target plane area while maintaining a certain positional relationship with respect to the target plane area.
The length of the linear light source is larger than the length of the target plane region so that an angle formed by two line-of-sight lines connecting both ends of the linear light source and both ends of the target plane region is a predetermined angle or more. It is set long,
Furthermore, the linear light source has a light direction restricting means for restricting an emission direction of the irradiation light along a plane including the two lines of sight to the range of the predetermined angle,
Thereby, the incident light amount of the illumination light is constant at every point in the target plane region, and the distribution range of the incident angle of the illumination light along the plane is constant within the range of the predetermined angle. The three-dimensional object image scanner according to any one of claims 1 to 4 . When the target plane area is viewed from the telecentric imaging system, one of the two light sources illuminates the target plane area from substantially the front, and the other light source illuminates the target plane area from an oblique direction. The three-dimensional object image scanner according to any one of claims 1 to 5, wherein The first and second light sources of the three-dimensional object image scanner according to one of the output light amount claim adapted to be independently variably set respectively 1-6. Assuming an XYZ orthogonal three-dimensional coordinate system in which the observation direction is the Z direction, the target plane region is a linear region of a predetermined length parallel to the Y direction perpendicular to the observation direction,
The moving means divides the subject space into a plurality of layers having different positions in the observation direction, and each of the plurality of layers has a different position in the Y direction and a width in the Y direction. The target plane area is divided into a plurality of bands set equal to the length of the target plane area, and the target plane area is arranged at the position of each layer in the observation direction by moving the target plane area in the observation direction. The target plane area is arranged at the position of each band in the Y direction by moving the target plane area in the Y direction at the layer position, and the target plane area is positioned in the observation direction at the position of each band. And the image data of each band is output from the image sensor by moving in the X direction perpendicular to the Y direction,
The image processing unit generates the image data of the plurality of layers by collecting the image data of the plurality of bands output from the image sensor, and the generated image data of the plurality of layers is set as the X direction. The pixel data in focus on the subject is extracted from the image data of the plurality of layers by comparing for each coordinate in the Y direction.
The three-dimensional object image scanner according to any one of claims 1 to 7 .

Images