JPH0634238B2 - 3D image display device - Google Patents

Description

TECHNICAL FIELD The present invention relates to a three-dimensional image display device, and more particularly to a three-dimensional image display device according to a density gradient of each point of an image.

[Conventional technology]

The display of medical images by a computer has been developed from a conventional two-dimensional display method to a three-dimensional display method. For example, "Isotope News"
(December 1985 issue, published by Japan Isotope Association) "3D display of medical images by computer" (pages 8-9).

This conventional example discloses that there are a surface method and a voxel method as surface disclosure methods. In the surface method, the density is given in correspondence with the inclination angle of each point of the image, and the density is displayed on the screen to express the shadow.

In the voxel method, the density is given according to the depth distance from the viewpoint or the light source to each point on the image, and this density is displayed on the screen to express the shadow.

[Problems to be Solved by the Invention]

Although the voxel method is excellent in grasping the depth distance, it has a drawback that the shape of a fine portion is not accurately grasped as compared with the surface method. On the other hand, in the surface method, the shape of a fine portion can be accurately grasped, but the depth distance is not accurately grasped. Shading in the three-dimensional display of an image can be realized by using one of the voxel method and the surface method, but in order to bring out the respective features, the position and shape of the image, or the point of interest should be selected. It is preferable to use them properly.

Therefore, when displaying one screen, it is necessary to consider that one is the voxel method and the other is the surface method.

However, both the voxel method and the surface method have the above-mentioned drawbacks. Therefore, the pure voxel method and the pure surface method cannot achieve shading under a more accurate three-dimensional display.

It is an object of the present invention to provide a three-dimensional display device capable of more accurate shading by correcting density data obtained by the voxel method by the surface method.

[Means for Solving the Problems]

According to the present invention, the density obtained by the voxel method is corrected at each point by adding the density gradient of the density of the neighboring pixels to obtain the corrected density, and the corrected density is displayed on the display surface. .

[Action]

According to the configuration of the present invention, the corrected density data is calculated for each point and the three-dimensional shade display is performed.

〔Example〕

An embodiment of a processing apparatus to which the present invention is applied is shown in FIG.
The computer (CPU) 201 is the center of the processing device. The high-speed dedicated hardware (correction density calculation device) 202 calculates correction density data, which is the center of the processing of the present invention.
The disk or drum memory 203 is a large capacity memory and stores data before processing and data after processing. The display device 204 is, for example, a CRT display device and performs three-dimensional display. The above devices 201 to 204 are common bus 2
It is tied with 05.

The three-dimensional display target data may be CT image data or projection data.

For example, an example of CT slice three-dimensional display will be described with reference to FIG. A plurality of CT images # 1, # for the subject with a CT device
Get 2, # 3, ... For example, up to 10 in the human head
It will be about 0 sheets. By stacking a plurality of CT slice images thus obtained, a three-dimensional image is obtained.

There is a three-dimensional display method as this three-dimensional image display method. The three-dimensional display method is a method of producing an apparent three-dimensional image mainly by computer graphic technology and projecting and displaying this on a normal two-dimensional display.

The three-dimensional display method includes a cross-section conversion display method and a surface display method. The cross-section conversion display method is a method of displaying a human body image collected as a three-dimensional data structure, that is, a three-dimensional image on an image memory, cutting it from any direction, and reproducing and displaying the cut surface as a two-dimensional image. Yes, the basic is high-speed interpolation calculation of the gray value corresponding to the coordinates.

There is a surface display method as a surface display method for the cut surface obtained by the cross-section conversion display method, or a method for properly displaying the surface of the image itself irrespective of the cut surface.

This surface display method is a method of forming a three-dimensional image of the surface shape of a complicated intricate organ, bone, or the like. Typical surface display methods include the surface method and the voxel method. The two methods have been described above. The voxel method is also called a painting algorithm method.

First, it is assumed that the disk memory 203 stores a three-dimensional voxel image obtained by the voxel method arranged for three-dimensional display. The computer 201 is used to obtain a voxel image from the original image by the voxel method. Calculator 20
In No. 1, the pixel density data of the original image stored in the disk memory 203 one after another is taken out, and the processing by the voxel method is performed to obtain the three-dimensional voxel pixel density data. This three-dimensional voxel pixel density data is again stored in the disk memory 20.
Store in 3. This is shown in FIG. A state in which the voxel pixel density data A _ij is stored for each pixel is shown. Note that n indicates the size of the screen or the size of the screen to be processed by the voxel method.

The corrected density data according to the present invention is the value B _ij calculated by the density gradient near the voxel pixel position (i, j). This modified density data B _ij is used as the voxel pixel density A _ij.
Replace on behalf of. Since the corrected density data B _ij is the value obtained from the density gradient, it is the density itself corrected by the surface method. This calculation work is performed by the high-speed dedicated hardware 202.

The procedure is shown in FIG. FIG. 4 shows an explanatory view thereof.

The data before processing is the voxel pixel density data shown in FIG. 2, and this data may remain stored in the disk memory 203 or may be brought to another high-speed memory. May be. In the following, it is assumed that the processing is performed by the high-speed dedicated hardware after the transfer to the high-speed memory.

In FIG. 3, steps 10 and 100 are processes by the computer 201, and steps 20 to 90 are processes by the high-speed dedicated hardware 202.

Further, FIG. 4 shows a state in which the display example based on the voxel pixel density data is converted into the display example based on the corrected density data. The original image is a voxel image. Density correction by neighboring pixels is performed for all pixels of this original image. In FIG. 4, assuming that the density at the scanning pixel is A, the density A is calculated by the pixels A ₁ , A ₂ , A ₃ , and A ₄ at four points that are one pixel in the peripheral biaxial direction through the scanning pixel. An example corrected to B is shown. This correction is performed for all n × n pixels.

Now, the processing procedure will be described in detail with reference to FIG.

In step 10, the original image data in the disk memory 203 is transferred to the main memory (not shown) in the computer 201. After this movement, the computer 201 executes processing on the dedicated hardware 2
Leave it to 02.

First, the dedicated hardware 202 executes steps 20, 30.
Initialize the access address of the original image with. If the access method based on the raster scan method is adopted, it means that the access method is based on (x ₁ , y ₁ ) in FIG.

For this (x ₁ , y ₁ ), neighboring pixels are extracted as shown in FIG. 4 (b) (step 40). In the figure,
The density according to (x ₁ , y ₁ ) is A, and the pixel densities on both sides x ₁ −1 and x ₁ +1 in the x direction are A ₂ , A ₄ , and y ₁ −1, y ₁ +1 in the y direction. The pixel densities are A ₁ and A ₃ . The pixel density gradients (A ₂ −A ₄ ) and (A ₁
-A ₃ ) is taken into consideration to obtain the corrected density data B at (x ₁ , y ₁ ) (step 40).

The general formula of the calculation formula of the corrected density data B is as follows.

B = | cos θ ₁ × cos θ ₂ | ^δ (1) δ is a constant determined by the reflection state of the original object from which the image was obtained. Δ = 1 for a non-reflector and δ> for a reflector.
It becomes 1. Therefore, the closer to the mirror surface, the larger the value of δ. For example, it is usual to set δ = 1.5 to 2.0 for metal, δ = 1 for gypsum, and δ = 1 for human body (however, δ may be changed depending on bone and skin). Angle θ ₁
Is the tilt angle (tilt) of a tilted line (a straight line connecting A ₂ and A ₄ ) with respect to the x-axis, and the angle θ ₂ is the tilt of a tilted line (a straight line that connects A ₁ and A ₃ ) with respect to the y-axis. The angle (tilt). cos θ ₁ indicates a component of the angle θ ₁ parallel to the x-axis, and cos θ ₂ indicates a component parallel to the y-axis.

On the other hand, the equation (1) can be replaced with the following equation in the neighborhood four-point method as shown in FIG.

Here, K represents a contrast constant, and G represents a distance per unit density. For example, K = n and G = 1.

Thus, B obtained by the equation (1) or (2) is (x ₁ ,
It becomes a corrected density corrected by the surface method in y ₁ ) and replaces the density A obtained by the voxel method (step 50).

In step 60, one point of interest is moved to the right with x _{1 set} to (x ₁ +1), and in step 70, it is checked whether or not x _max is reached. In step 80, y _{1 is changed} to (y ₁
+1), one point is moved downward, and in step 90, it is checked whether or not y _max is reached. These steps 60 to 90 merely show scanning by the raster scanning method.

Step 100 is processing by the computer 201, which stores the result and displays it on the display device 204. Results are shown in FIG. This storage is finally stored in the disk memory 203.
When displaying, it is stored in the refresh memory and then displayed. It should be noted that although in the embodiment, the perimeter is set to one point, the perimeter can be accurately detected if the perimeter is set to two points, the perimeter to three points, and the like, and more accurate correction can be performed.

Although the method of correcting the pixel density data of the voxel method is adopted, both the density data of the voxel method and the corrected density data may be used. In that case, the density data of the voxel method may be left as it is, and the modified density data may be calculated based on the density data of the voxel method and stored together in the memory.

Further, although the inclinations are separate in the x direction and the y direction, they may be entangled and corrected. Further, it can be realized by an arithmetic unit processed by a microprogram instead of the high-speed dedicated hardware.

〔The invention's effect〕

According to the present invention, pixel density data by the voxel method
It can be corrected by the surface method.

[Brief description of drawings]

FIG. 1 is an embodiment diagram of the present invention, FIG. 2 is an example diagram of concentration data of the voxel method, FIG. 3 is a diagram showing a processing procedure of the present invention, and FIG.
FIG. 5 is an explanatory diagram of an image example, FIG. 5 is an exemplary diagram of corrected density data, and FIG. 6 is an explanatory diagram of a three-dimensional display image. 201 ... Computer (CPU), 202 ... High-speed dedicated hardware, 203 ... Disk memory, 204 ... Display device.

Claims (2)

[Claims] 1. A memory for storing a three-dimensional voxel image,
The three-dimensional voxel image is composed of a processing means for obtaining a corrected density corrected by the surface method from each pixel density, and a display means for displaying the three-dimensional image of the corrected density. Gradients θ ₁ and θ ₂ of voxel pixel densities of two intersecting axes (where θ ₁ is the inclination of one of the two axes and θ ₂ is the inclination of the other of the two axes).
And the pixel density B = | cos θ for each pixel from this slope
₁ × cos θ ₂ | ^δ (where δ is a constant determined by the reflection state of the original object from which the image was obtained), and the calculated pixel density B is the corrected density corrected by the surface method. 3D image display device. 2. A memory for storing a three-dimensional voxel image,
The processing means includes a processing means for obtaining a corrected density corrected by the surface method from each pixel density of the three-dimensional voxel image, and a display means for displaying the three-dimensional image of the corrected density, and the processing means includes each voxel pixel. For each of the two axes of the voxel pixel on two axes intersecting the voxel pixel, the pixel density (A ₁ ,
_{A 3), (A 2,} A 4) the slope of _{(A 2 -A 4), (} A 1
-A ₃ ) is calculated, the pixel density B is calculated from this slope by the following equation, (However, K is the contrast constant, G is the distance per unit density, and δ is a constant determined by the reflection state of the original object from which the image was obtained.) This calculated pixel density B was corrected density by the surface method. A three-dimensional image display device.