JPH09237352A - Three-dimensional picture constituting method and device therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To visualize the internal shape of a specimen in a wide range by shading corresponding to the internal shape of the specimen with respect to a parallelly projected two-dimensional picture. SOLUTION: In this method, a specimen with an internal space among volume pictures expressed by a world coordinate system constituted by piling many tomographic images 30 is perspectively converted into a coordinate system at an optional view point face 20 to obtain a two-dimensional picture, and the two dimensional picture is shaded based on a distance NRD between an outline point outside of the specimen and the view point face to obtain a three- dimensional picture. A virtual line light source is provided inside of this specimen to obtain a distance between the position of a light source and an outline point inside of the specimen on each tomographic image. Then the distance is reflected to the shading to the two dimensional picture to shade the two-dimensional picture of the specimen corresponding to its surface shape and inside shape. Thereby, as the outside surface of the specimen is given the shading of the inside of the specimen, internal unevenness can be confirmed and internal information is obtained by looking at its outside.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for constructing a three-dimensional image, and in particular, for viewing a volume image obtained by stacking a large number of tomographic images obtained from an X-ray CT apparatus, an MRI apparatus or the like from any direction. The perspective plane is transformed into a two-dimensional image, and the two-dimensional image is shaded to give a three-dimensional image. (The three-dimensional image here is a three-dimensional image in which each pixel of the two-dimensional array is given a density. A three-dimensional image composing method and apparatus.

[0002]

2. Description of the Related Art As a three-dimensional image construction method of this type,
As shown in FIG. 10, a viewpoint plane 20 corresponding to a display screen for viewing the inspection object is appropriately set, and each tomographic image 3 of the inspection object is set.
A method of extracting a contour point illuminated by a virtual parallel light source from the viewpoint plane 20 with respect to 0 and performing shading on a two-dimensional image of the object to be inspected, which is a set of the extracted contour points, according to a shading algorithm. There is.

That is, the contour points of the object to be inspected which are visible from the viewpoint plane 20 are extracted, and the contour points are perspective-transformed into the coordinate system of the viewpoint plane 20, and the brightness of the perspective-transformed contour points is changed to the contour points and the viewpoint plane. The decision is made based on the distance from. For example, the contour points near the viewpoint plane are brightened, and the contour points far from the viewpoint plane are darkened.

[0004]

However, in the above-described three-dimensional image construction method, since the parallel light source is located outside the object to be inspected, the target organ such as the trachea, esophagus, stomach, or intestine is located outside the object. Although it is effective in constructing it as a three-dimensional image viewed from above, there was a problem that information inside it could not be obtained. That is, it was not possible to obtain information on the protrusion (that is, the lesion) surrounded by the circle in FIG.

On the other hand, for example, Japanese Laid-Open Patent Publication No. 4-13683.
In the three-dimensional image construction method described in Japanese Patent Laid-Open No. 1, a viewpoint is placed inside an object to be inspected, coordinate conversion by central projection is used to project each tomographic image onto a projection surface, and the object to be inspected is viewed from the inside. I am trying to compose an image. However, the image formed by this method becomes an image of an object to be inspected by an endoscope, and there is a problem that the observation range is narrow.

The present invention has been made in view of the above circumstances, and a two-dimensional image projected in parallel can be shaded according to the internal shape of the object to be inspected. An object of the present invention is to provide a three-dimensional image constructing method and apparatus which can be visually recognized over a wide range.

[0007]

In order to achieve the above-mentioned object, the present invention provides an arbitrary viewpoint of an object to be inspected having an internal space among volume images represented by a world coordinate system in which a large number of tomographic images are stacked. A two-dimensional image is obtained by perspective transformation into a coordinate system on a plane, and the two-dimensional image is shaded based on the distance between the contour point outside the inspection object and the viewpoint plane to form a three-dimensional image. In the three-dimensional image forming method, the contour points inside the inspection object are extracted for each tomographic image, a virtual linear light source is provided inside the inspection object, and the position of the light source on each tomographic image is It is characterized in that a distance from the inner contour point is calculated, and shading is further performed on the two-dimensional image of the inspection subject based on the calculated distance. That is, when performing the perspective transformation and the shading in the three-dimensional image construction method, the perspective transformation is performed by a conventional method, and the shading is performed by the distance from the virtual light source inside the inspected object for each tomographic image. It is applied to the outside of the inspection object according to the calculated shading algorithm. As a result, when a three-dimensional image of the inspection object is constructed, the outer surface of the inspection object is shaded inside the inspection object, so that it is possible to confirm the internal unevenness by looking at the outer surface, Inside information is obtained.

According to another embodiment of the present invention,
A volume image represented by a world coordinate system, which is a stack of multiple tomographic images, is perspective-transformed into a coordinate system at an arbitrary viewpoint plane to obtain a two-dimensional image, and the two-dimensional image is shaded to obtain a three-dimensional image. In the three-dimensional image configuration method configured as an image, to extract the contour points inside the inspection object having an internal space for each tomographic image, and perspectively transform the extracted contour points into the coordinate system of the viewpoint plane, The distance between the contour point in the world coordinate system and the viewpoint plane is calculated, and the contour point farther from the viewpoint plane is selected from the two contour points at the same position in the coordinate system on the viewpoint plane, and the selected contour is selected. The feature is that the density is added to the point according to the distance from the viewpoint surface, and the inner two-dimensional image obtained by cutting the inspection object is shaded. That is, the object to be inspected is cut by a plane passing through its center, the cut inner contour points are perspective-transformed into a viewpoint plane, and shading is performed according to a shading algorithm. As a result, it is possible to confirm the cut inside unevenness of the inspection object.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of a three-dimensional image construction method and apparatus according to the present invention will be described in detail below with reference to the accompanying drawings. First, the principle of the three-dimensional image forming method according to the present invention will be described with reference to FIGS. As shown in FIG. 1, for example, the stomach is taken out as a target organ from each tomographic image 30 of the inspected object which is perspective-transformed to the viewpoint 20.
In this case, only the stomach can be taken out by ignoring pixels of organs other than the stomach and performing perspective transformation.

A virtual linear light source C penetrating the inside of the stomach (inspection object) is set. This line light source becomes a point light source on each tomographic image 30. The position of this linear light source may be set, for example, by calculation so as to penetrate the position of the center of gravity inside each tomographic image 30, or may be set to an approximate position of the center of gravity inside each tomographic image 30 by operating means. May be.

Then, the shading calculation is performed on the pixels inside the target organ illuminated by the light source set inside the target organ, and the internal information can be confirmed from the result of the shading calculation. This will be described with reference to FIGS. Here, the range surrounded by the curve 21 on the tomographic image 30 represented by the world coordinate system (x, y, z) is the target organ, and the position of the point light source arbitrarily given inside the target organ on the operating side. Be c (x0, y0, z0). Also, the position c
To r, the increment of the distance r is Δr, the direction from the position c is θ, the contour point inside the organ is p1 (x1, y1, z1),
The outer contour point is p2 (x2, y2, z2). In addition, p1,
p2 is given by c, r and θ.

When the contour point p1 inside the target organ is extracted as shown in FIG. 3, the distance r at that time is stored as the distance R from the light source c. When an outer contour point p2 in the same direction as the contour point p1 is extracted, the contour point p2 and the viewpoint plane 2 are extracted.
A distance NRD from 0 and a point p'2 on the viewpoint plane 20 of the contour point p2 are obtained. Here, the distance NRD between the contour point p2 and the viewpoint 20 is obtained as follows.

The origin O in the world coordinate system is set to the viewpoint 2
The point projected on 0 is O ′, and the distance of the line segment OO ′ is NE.
Let Z be the angle that the line segment OO ′ makes with the xz plane, α be the angle that the projection of the line segment OO ′ makes with the xz plane be the β axis, and let y2 be the tomographic image at which the point p2 exists. The distance NRD from the surface 20 is calculated by the following equation:

[0014]

## EQU1 ## NRD = NEZ-x2.cos .beta.cos .alpha.-y2.cos .beta.sin .alpha.-z2.sin .beta. (1) In addition, the point p2 projected on the viewpoint plane 20
The coordinates of are in the world coordinate system.
To convert to the XY coordinate system at 0, the following formula is used.

[0015]

[Formula 2] X = x2 · cos α−z2 · sin α (2)

[0016]

## EQU00003 ## Y = x2.sin .alpha.sin .beta. + Y2.cos .beta. + Z2 .cos .alpha.sin .beta. (3) On the other hand, the point p2 perspective-transformed into the XY coordinate system of the viewpoint plane 20
The concentration value of is calculated by the following formula.

[0017]

## EQU00004 ## Density value of viewpoint plane = constant.R.NRD (4) When the above processing is performed for all target organ regions of the tomographic image calculated as shown in FIG. 4, the XY coordinate system of the viewpoint plane 20 is obtained. Although the two contour points overlap with each other at the same position, the contour point of the object to be inspected seen from the viewpoint plane 20 is the contour point closer to the viewpoint plane 20, and therefore the contour point closer to the viewpoint plane 20 is Select data. The light source is a curved light source that passes through the center of gravity inside the target organ, and the density value for shading including the internal information of the target organ is calculated. Then, by using this density value for shading the outside of the target organ, it is possible to construct a three-dimensional image in which information on the inside of the organ as well as the outside of the target organ can be obtained. That is, in this constituent image, the unevenness P0 on the inner surface of the target organ is displayed as a shadow P0 'on the surface of the target organ.

FIG. 5 is a block diagram showing a hardware configuration to which the present invention is applied. In the figure, 50 is a central processing unit (CPU), 51 is a main memory, 52 is a magnetic disk, 53 is a display memory, 55 is a mouse controller, and these are connected to a common bus 58. The magnetic disk 52 stores a volume image represented by a world coordinate system in which a large number of tomographic images are stacked, a program for coordinate conversion and shading, and the like.

The CPU 50 reads out a large number of tomographic images and programs for coordinate conversion and shading as will be described later, and the main memory 51 is used to shade a three-dimensional image of the inside shape of the object to be inspected. An image is constructed, and the result is sent to the display memory 53 and displayed on the CRT 54. A mouse 56 and a keyboard 57 connected to the mouse controller 55 specify a viewpoint surface, a light source position, and the like when forming a three-dimensional image. The obtained three-dimensional image is stored in the magnetic disk 52 as needed.

Next, the processing contents of the CPU 50 will be described with reference to the flowcharts shown in FIGS. First, a sufficiently large value is set in the Z buffer memory that stores the distance value at the time of initial setting, and the viewpoint plane, the target organ region, the threshold value for contour extraction, etc. are set (step S10). Then, the tomographic image corresponding to the predetermined position (y0) is read from the magnetic disk 52, and the reference value c (x0, y0, z0) that is the position of the light source with respect to the read tomographic image is obtained (step S12). In this case, the barycentric position of the read tomographic image is calculated, and the barycentric position is used as the reference position.

Next, the angle θ = 0 and the distance r = 0 are initialized (step S14), and the point p1 is calculated based on the reference values c and θ, r as follows:

[0022]

## EQU00005 ## p1 (x1, y1, z1) = (x0 + r.cos .theta., Y0, z0 + r.sin .theta.) (5) is calculated (step S16). It is determined whether or not the pixel value of the point p1 (CT value in the case of a tomographic image obtained by an X-ray CT apparatus) thus obtained satisfies the threshold value as an inner contour point of the object to be inspected ( Step S18).

If the pixel value at the point p1 does not satisfy the threshold condition, the distance r is increased by a minute amount Δr and set as a new distance r (step S20). If the distance r is not the maximum value (in the target organ area), the process returns to step S16, and until the pixel value of the point p1 satisfies the threshold condition or the distance r reaches the maximum value, step S16. The process of step S22 is executed.

On the other hand, when the pixel value of the point p1 satisfies the threshold condition, the distance r at that time is stored as the distance R from the reference value c to the contour point inside the object to be inspected (step S
24) Further, the distance NRD between the point p1 and the viewpoint plane is obtained (step S26). It is determined whether or not this distance NRD is smaller than the initially set Z buffer value (step S28), and if it is smaller, this distance NRD is stored in the Z buffer memory (step S30).

Next, as shown in FIG. 7, the distance r is increased by a minute amount Δr and set as a new distance r (step S32), and the point p2 is determined based on the reference values c and θ, r.
The following formula,

[0026]

[Mathematical formula-see original document] p2 (x2, y2, z2) = (x0 + r * cos [theta], y0, z0 + r * sin [theta]) (6) (step S34). It is determined whether or not the pixel value of the point p2 thus obtained satisfies the condition as the contour point outside the inspection object (step S36).
In this case, it is determined whether or not the pixel value of the point p2 first deviates from the threshold value. When the pixel value of the point p2 does not satisfy the condition as the outer contour point, the distance (NRD
−Δr · cos θ) is calculated as a new distance NRD (step S38), the process returns to steps S28 and S30, and the contents of the distance NRD stored in the Z buffer memory are rewritten.

On the other hand, when it is determined in step S36 that the pixel value of the point p2 satisfies the condition as the outer contour point, the density value of the contour point on the viewpoint surface (the density of one pixel is the number of dots If it is displayed, the number of dots)
Is calculated by the equation (4) described above (step S4
0). In addition, the position of the point p2 in the world coordinate system is
Convert to the XY coordinate system of the viewpoint plane (Equation (2) and (3) above)
Expression), and the calculated density value and distance NRD corresponding to the coordinate position are stored (step S42).

Next, the angle θ is increased by a minute angle Δθ (θ = θ + Δθ), the distance r is set to 0 (step S44), and it is determined whether θ> 2π (step S4).
6). If θ> 2π is not satisfied, step S16 in FIG.
Return to That is, the scanning direction from the reference point c is changed, the contour points inside and outside the object to be inspected are extracted again, and the density value of the contour points is calculated and the coordinates are converted.

On the other hand, when θ> 2π, for one tomographic image, all contour points inside and outside the object to be inspected in the tomographic image are extracted, density values are calculated, coordinate conversion is performed. It has been broken. Therefore, in this case, it is determined whether or not the processing of all tomographic images of the subject (target organ region) has been completed (step S48). If processing of all tomographic images is not completed, the process returns to step S12 of FIG.
The next tomographic image is read from the magnetic disk 52 and the same processing as above is executed.

When the processing for all tomographic images of the target organ region is completed in this way, the process proceeds to step S50.
In this step S50, since there are two contour points that have been coordinate-converted into the XY coordinate system of the viewpoint plane at the same coordinate position in the XY coordinate system, they are captured corresponding to one of the contour points closer to the viewpoint plane. The density value (density value of the contour point having the smaller NRD) is adopted and stored in the display memory 53.
As a result, the construction of the three-dimensional image for obtaining the information inside the organ while extracting the outside of the target organ is completed. The three-dimensional image thus configured generates a luminance signal based on the density value of each pixel on the viewpoint stored in the display memory 53, and outputs the luminance signal to the CRT 54,
It is possible to display an image on the surface of the target organ, on which information such as irregularities inside the target organ is reflected, on the screen of the CRT 54. It is also possible to obtain a hard copy of the image by connecting the display memory 53 to a hard copy device.

Next, another embodiment of the present invention will be described with reference to the flowcharts of FIGS. 8 and 9.
Note that the same parts as those in the flowcharts shown in FIGS. 6 and 7 are designated by the same reference numerals. In this embodiment, the light source is not provided inside the object to be inspected, the object to be inspected is cut by a plane parallel to the viewpoint surface, and the contour points inside the object are perspective-transformed to the viewpoint surface and the shading algorithm is followed. It is shaded so that the inside of the object can be checked directly.

First, after initializing the Z buffer value, the viewpoint plane, etc. (step S10), a tomographic image corresponding to a predetermined position (y0) is read from the magnetic disk 52, and the reference value c for the read tomographic image is read. (X0, y0,
z0) is calculated (step S12). Reference value c in this case
Is the reference position of the scan for extracting the contour line inside the tomographic image, and may be inside the tomographic image and may not be at the center of gravity of the tomographic image. Then, angle θ = 0, distance r = 0
(Step S14), and the point p1 is calculated based on the reference values c and θ, r (step S16). Then, it is determined whether or not the pixel value of the point p1 thus obtained satisfies the threshold condition while changing the distance r, and thereby the contour point inside the inspection object is extracted (step S1).
6 to step S22).

If the pixel value of the point p1 satisfies the threshold condition, the distance NRD between the point p1 and the viewpoint plane is obtained (step S26), and this distance NRD is set to the initially set Z.
If it is smaller than the buffer value, it is stored in the Z buffer memory (steps S28 and S30).
Next, the density value of the point p1 extracted as the contour point on the viewpoint plane is calculated from the distance NRD between the point p1 and the viewpoint plane (step S100). Further, the position of the point p1 in the world coordinate system is converted into the XY coordinate system of the viewpoint plane, and the calculated density value and distance NRD corresponding to the coordinate position are stored (step S42).

Next, the angle θ is increased by a minute angle Δθ and the distance r is set to 0 (step S44), θ>
It is determined whether it is 2π (step S46). If θ> 2π is not true, the process returns to step S16 in FIG. That is, the scanning direction from the reference point c is changed, the contour points inside the object to be inspected are extracted again, and the density value of the contour points is calculated and the coordinates are converted.

On the other hand, in the case of θ> 2π, for one tomographic image, all contour points inside the object to be inspected in the tomographic image are extracted, density values are calculated, and coordinate conversion is performed. It will be. Therefore, in this case, it is determined whether or not the processing of all the tomographic images of the inspection object is completed (step S4).
8) If processing of all tomographic images is not completed,
Returning to step S12, the next tomographic image is recorded on the magnetic disk 5
2 is read out, and the same processing as above is executed.

When the processing for all the tomographic images of the target organ region is completed in this way, the process proceeds to step S102. In this step S102, since there are two contour points coordinate-converted into the XY coordinate system of the viewpoint plane at the same coordinate position in the XY coordinate system, the contour points farther from the viewpoint plane are captured. The density value (density value of the contour point having the larger NRD) is adopted and stored in the display memory 53. As a result, the target organ is cut along the plane passing through the center, the inside of the cut target organ is perspective-transformed, and the construction of the three-dimensional image is completed.

In this embodiment, the contour of the entire circumference of the object to be inspected is extracted and then the contour to be actually shaded is selected. However, the present invention is not limited to this, and it is based on the setting of the viewpoint plane. Alternatively, the contour extraction range (angle θ shown in FIG. 3) may be determined. In step S100 described above, the distance r between the light source and the contour point may be added as a parameter when determining the density value on the viewpoint plane of the point p1 extracted as the contour point.

[0038]

As described above, according to the present invention, the two-dimensional image projected in parallel is shaded according to the internal shape of the object to be inspected. Therefore, it is originally obtained from the outside of the object to be inspected. It is possible to confirm the information (shape) of the inside of the inspected object that cannot be seen, and because it is a two-dimensional image that is projected in parallel, the inside of the inspected object can be observed compared to the image seen with an endoscope. The range can be expanded.

[Brief description of drawings]

FIG. 1 is a diagram used for explaining the principle of the present invention.

FIG. 2 is a diagram showing a tomographic image represented by a world coordinate system and a viewpoint plane used for explaining the principle of the present invention.

FIG. 3 is a diagram used for explaining contour extraction on the inside and outside of a test object having an internal space.

FIG. 4 is a diagram showing a manner in which a three-dimensional image is constructed by perspective-transforming an object to be inspected into a viewpoint plane according to the present invention.

FIG. 5 is a block diagram showing a hardware configuration to which the present invention is applied.

FIG. 6 is a flowchart used to explain the processing contents of the CPU shown in FIG.

FIG. 7 is a flowchart used to explain the processing contents of the CPU shown in FIG.

FIG. 8 is a flowchart used to explain another embodiment of the processing contents of the CPU shown in FIG.

9 is a flowchart used to describe another embodiment of the processing contents of the CPU shown in FIG.

FIG. 10 is a diagram used for explaining an example of a conventional three-dimensional image forming method.

[Explanation of symbols]

 20 ... View plane 30 ... Tomographic image 50 ... CPU 51 ... Main memory 52 ... Magnetic disk 53 ... Display memory 54 ... CRT 56 ... Mouse 57 ... Keyboard 57 c ... Reference value (light source position) NRD ... Contour point and view plane Distance R: Distance between light source and inner contour point

Claims (10)

[Claims] 1. A two-dimensional image is obtained by perspectively transforming an object to be inspected having an internal space into a coordinate system of an arbitrary viewpoint plane among volume images represented by a world coordinate system in which a large number of tomographic images are stacked. At the same time, in the three-dimensional image constructing method for constructing a three-dimensional image by shading the two-dimensional image based on the distance between the contour points outside the inspection object and the viewpoint plane, An outline point inside the inspection object is extracted, a virtual line light source is provided inside the inspection object, the distance between the position of the light source on each tomographic image and the inside outline point is calculated, and the calculation is performed. A method for constructing a three-dimensional image, wherein the two-dimensional image of the object to be inspected is further shaded based on the distance. 2. A volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, is perspective-transformed into a coordinate system on an arbitrary viewpoint plane to obtain a two-dimensional image, and the two-dimensional image is shaded. In the three-dimensional image constructing method of performing a three-dimensional image by performing the above, the inside contour points and the outside contour points of the inspected object having an internal space for each tomographic image are extracted, Line light source is provided, a first distance between the position of the light source on each tomographic image and the inner contour point is calculated, and the outer contour point of the extracted world coordinate system is set to the coordinate system on the viewpoint plane. And the second distance between the contour point outside the world coordinate system and the viewpoint plane is calculated, and one of the two contour points at the same position in the coordinate system on the viewpoint plane that is closer to the viewpoint plane is calculated. While selecting the contour points, By attaching a density corresponding to the first distance and the second distance to the selected contour point of, the two-dimensional image of the inspection object is shaded corresponding to the surface shape and the inner surface shape. A method for constructing a three-dimensional image characterized by the above. 3. A volume image represented by a world coordinate system in which a large number of tomographic images are stacked is perspectively transformed into a coordinate system on an arbitrary viewpoint plane to obtain a two-dimensional image, and the two-dimensional image is shaded. In the three-dimensional image constructing method for constructing as a three-dimensional image, the outer contour points and inner contour points visible from the viewpoint plane of the inspection object having an internal space for each tomographic image are extracted, A virtual line light source is provided inside the inspection body, a first distance between the position of the light source on each tomographic image and the inner contour point is calculated, and the outer contour point of the extracted world coordinate system is defined as The second distance between the contour point outside the world coordinate system and the viewpoint plane is calculated while performing the perspective transformation into the coordinate system on the viewpoint plane, and the first distance and the second distance with respect to the contour point subjected to the perspective transformation. The concentration corresponding to the distance of Three-dimensional image constructing method being characterized in that to perform the shading corresponding to the surface shape and the inner shape in a two-dimensional image of the object to be inspected by kicking. 4. A volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, is perspective-transformed into a coordinate system on an arbitrary viewpoint plane to obtain a two-dimensional image, and the two-dimensional image is shaded. In the three-dimensional image construction method of performing as a three-dimensional image, to extract the contour points inside the inspection object having an internal space for each tomographic image, the extracted contour points in the coordinate system of the viewpoint plane While performing perspective transformation, the distance between the contour point in the world coordinate system and the viewpoint plane is calculated, and the contour point farther from the viewpoint plane is selected from the two contour points at the same position in the coordinate system on the viewpoint plane. ,
A three-dimensional image constructing method characterized in that the selected contour point is shaded on the inner two-dimensional image obtained by cutting the object to be inspected by giving a density corresponding to the distance from the viewpoint plane. . 5. A two-dimensional image is obtained by perspective-transforming a volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, into a coordinate system on an arbitrary viewpoint plane, and is shaded on the two-dimensional image. In the three-dimensional image constructing method for constructing a three-dimensional image by performing the following, the contour points on the inside of the inspected object having an internal space for each tomographic image, from the viewpoint plane when the inspected object is cut. The visible contour points are extracted, the contour points of the extracted world coordinate system are perspective-transformed into the coordinate system of the viewpoint plane, and the distance between the contour points of the world coordinate system and the viewpoint plane is calculated, and the perspective transformation is performed. A method for constructing a three-dimensional image, characterized in that a density corresponding to a distance from a viewpoint plane is added to a contour point so that an inner two-dimensional image obtained by cutting an object to be inspected is shaded. 6. The contour points are extracted by scanning a tomographic image while changing a distance and a direction from a reference point inside an object to be inspected, and extracting the contour points based on pixel values indicating the scanned tomographic image. The three-dimensional image construction method according to claim 1, wherein 7. A two-dimensional image is obtained by perspective-transforming a volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, into a coordinate system on an arbitrary viewpoint plane, and is shaded on the two-dimensional image. In the three-dimensional image forming apparatus configured to perform a three-dimensional image by performing the above, a contour extracting unit that extracts an inner contour point and an outer contour point of the inspected object having an internal space for each tomographic image, and the inspected object A virtual linear light source inside, and a first calculation means for calculating a first distance between the position of the light source on each tomographic image and the inside contour point; and an outside of the extracted world coordinate system. Second computing means for calculating a second distance between a contour point and the viewpoint plane, and a contour point closer to the viewpoint plane among two contour points at the same position in the coordinate system on the viewpoint plane is selected. Selecting means, and the selecting means Density value determining means for determining a density value for the contour point based on the first distance and the second distance calculated corresponding to the selected contour point, and the density value determining means. A three-dimensional image forming apparatus comprising: a display storage unit that stores a density value as image information corresponding to each coordinate position on the viewpoint surface. 8. A two-dimensional image is obtained by perspective-transforming a volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, into a coordinate system on an arbitrary viewpoint plane, and is shaded on the two-dimensional image. In the three-dimensional image constructing device for performing the above-mentioned processing to construct a three-dimensional image, a contour extracting means for extracting an outer contour point and an inner contour point visible from the viewpoint plane of the object to be inspected having an internal space for each tomographic image. And a first calculation means for providing a virtual line light source inside the object to be inspected and calculating a first distance between the position of the light source on each tomographic image and the inside contour point, and the extraction. Second calculating means for calculating a second distance between the contour point outside the world coordinate system and the viewpoint plane, and based on the first distance and the second distance calculated corresponding to the contour point Determine the density value for the contour point And a display storage unit that stores the density value determined by the density value determination unit as image information in correspondence with each coordinate position on the viewpoint plane. Image composition device. 9. A volume image represented by a world coordinate system, which is formed by stacking a large number of tomographic images, is perspective-transformed into a coordinate system on an arbitrary viewpoint plane to obtain a two-dimensional image, and the two-dimensional image is shaded. In a three-dimensional image forming apparatus configured to perform a three-dimensional image by performing the above, a contour extracting means for extracting a contour point inside the inspected object having an internal space for each tomographic image, and the contour point of the extracted world coordinate system A first calculating means for calculating a distance between the viewpoint surface and the viewpoint surface; a selecting means for selecting a contour point farther from the viewpoint surface from two contour points at the same position of the coordinate system on the viewpoint surface; Density value determining means for determining a density value for the contour point based on the distance from the viewpoint plane calculated corresponding to the contour point selected by the selecting means; and the density value determined by the density value determining means. Three-dimensional image constructing apparatus comprising: the display memory means for storing in correspondence with the coordinate position of the point plane values as image information. 10. A two-dimensional image is obtained by perspectively transforming a volume image represented by a world coordinate system formed by stacking a large number of tomographic images into a coordinate system on an arbitrary viewpoint plane,
In a three-dimensional image constructing device for forming a three-dimensional image by shading the two-dimensional image, a contour point inside the inspection object having an internal space for each tomographic image, and cutting the inspection object Contour extraction means for extracting a contour point visible from the viewpoint surface, a calculation means for calculating a distance between the extracted contour point in the world coordinate system and the viewpoint surface, and a calculation corresponding to the contour point. The density value determining means for determining the density value for the contour point based on the distance from the viewpoint surface, and the density value determined by the density value determining means as image information corresponding to each coordinate position of the viewpoint surface. A three-dimensional image composing device, comprising: