JPS62186374A - Boundary extracting device - Google Patents

Abstract

PURPOSE:To extract a boundary of a 3-dimension binary stereoscopic image at a high speed by providing a 3-dimension stereoscopic image input means, a stereoscopic binary-coding means, a stereoscopic image turning means, a circumscribed rectangular prism extracting means and an AND operation means. CONSTITUTION:The 3-dimension stereoscopic image input means 11, a 3-dimension stereoscopic binary-coding means 12 binary-coding the 3-dimension stereoscopic image inputted by the means 11 at a prescribed threshold value, and a 3-dimension stereoscopic image turning means 13 turning the 3-dimension binary coding stereoscopic image obtained by the means 12 by a prescribed angle theta with respect to an x, y plane and turning the image by a prescribed angle phi with respect to a z axis, are provided. The circumscribed rectangular prisms of the 3-dimension binary-coding image turned by the means 13 is obtained by a circumscribed rectangular prism extracting means 14, the circumscribed rectangular prisms at respective angles theta (0-90 deg.) and phi (0-90 deg.) obtained by the means 14 are ANDed and an AND operation means 15 extracts the boundary of the 3-dimension stereoscopic image. The 3-dimension stereoscopic image boundary data obtained by the AND operation means 15 is stored in a memory and displayed on a display device.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides a boundary extraction device suitable for grasping the appearance and extent of concavities in a three-dimensional three-dimensional image, and for grasping various properties of a three-dimensional solid. Regarding.

[Technical background of the invention and its problems] Conventionally, a contour extraction device for three-dimensional stereoscopic images has not been particularly considered;
A contour extraction device for a two-dimensional image converts image data detected by an image detection device such as a television camera or CCO line sensor into 21I, stores this binary image pixel by pixel in a table memory, etc. There is a device that extracts the outline of an image by tracing the outline of an image from table image data and, if there is a recessed valve in the outline of the image, connecting the recessed valve with a straight line. .

Another example is to trace the outline of the image from the table image data, and if there is a concave part in the outline of the image, replace the concave part with a convex curve, or with a straight line that eliminates the concave part. There are also devices that extract contours.

However, since all of the contour extraction devices described above find contours by tracing the same values in a binary image, special software is required to track the image, and it takes a long time to extract the contours. There was a problem that it was necessary.

[Purpose of the invention]

The present invention was made based on these circumstances,
The purpose is to provide a boundary extraction device that can extract boundaries of a three-dimensional binary stereo image at high speed using simple image processing hardware.

[Essentials of the invention]

As shown in FIG. 1, the present invention includes a three-dimensional stereo image input means 11 for inputting a three-dimensional stereo image, and a three-dimensional stereo image input by the three-dimensional stereo image input means 11 at a predetermined threshold value. A three-dimensional solid image 21ia converting means 12 to be binarized, and this three-dimensional solid l! A three-dimensional stereo image obtained by rotating the three-dimensional binary stereo image obtained by the I-binarization means 12 by a predetermined angle θ with respect to the (x, y) plane and rotating it by a predetermined angle φ with respect to the Z axis (two). A rotating means 13, a circumscribed rectangular parallelepiped extracting means 14 for obtaining a circumscribed rectangular parallelepiped of the three-dimensional binary stereoscopic image rotated by the three-dimensional three-dimensional image rotating means 13, and each angle θ(0 ~90 degrees),
a logical product calculation means 15 for calculating the logical product of circumscribed rectangular parallelepipeds at φ (0 to 90 degrees) and extracting the boundary of the three-dimensional stereoscopic image; It stores data in memory and displays it on a display.

[Embodiments of the invention]

FIG. 2 is a system diagram showing the configuration of an embodiment of the present invention.

In the figure, a control computer 21c has a function of controlling each component of the apparatus according to a program and outputting commands for predetermined three-dimensional image processing. and an address data bus 23 are derived. A main storage section 24 and an auxiliary storage section 25 are connected to this address data bus 23, and each of these storage sections 24 and 25 stores a program for executing three-dimensional stereoscopic image processing in the control computer 21, and also stores necessary programs. It has the ability to temporarily store data. On the other hand, a three-dimensional stereoscopic image input device 26 is connected to the input/output bus 22. The three-dimensional stereoscopic image input device 26 uses, for example, an XI1CT scanner, an ultrasonic tomography device, or the like, and images a three-dimensional object as a subject and outputs analog image data.

In addition, 27 in the figure is a three-dimensional stereo image processing control unit, and based on the instructions from the control computer 21, the multipath 27 is
8 and the image data bus 29 as a binarization means @21iI! conversion unit 30, affine transformation unit 31 as a three-dimensional image rotation means, circumscribed rectangular parallelepiped extraction unit 32, logical product operation unit 33, three-dimensional image memory that stores intermediate results of each three-dimensional image processing and logical product data of the circumscribed rectangular parallelepiped at each angle. 34
, and a display 35 for outputting a stereoscopic image as a result of displaying the outline of the stereoscopic image.

The stereoscopic image binarization unit 30 includes a three-dimensional stereoscopic image input device 26
The analog image data output from the converter is converted into "1" and "○" by a predetermined threshold value and output. Also,
The affine transformation unit 31 converts the binary stereoscopic image or circumscribed rectangular parallelepiped input from the binarization rope 30 into an angle ±θ (0≦θ
≦90 degrees) and at an angle φ (0
φ≦90 degrees). The circumscribed rectangular parallelepiped extraction unit 32 extracts r in the binary stereoscopic image.
The minimum x, y, z coordinate values of the three-dimensional area of lJ, Xm1n,
Ymin, zln and maximum x, y, z coordinates 1i1Xm
Find ax, Ymax, Zmax, and find X 1n
≦X≦xmax, katsuy mtn≦y≦'y' max
, and 71n≦2≦zmax is set to “1”.

Next, we will explain the third part about the operation of this device configured in this way.
The explanation will be given with reference to the flowchart shown in the figure from time to time.

First, in step 1, a three-dimensional stereoscopic image of a subject (three-dimensional solid) is input as analog image data by a three-dimensional stereoscopic image input means.

This analog image data is processed by the three-dimensional stereoscopic image processing control section 27.
The 3D @binarization unit 30 converts the image into a binary image using a predetermined threshold value in step 2, for example, to create a 3D binary 3D image as shown in FIG. get. At this time, as shown in step 3, as an initial setting, both the stereoscopic image rotation angles θ and φ are set to 0 degrees, and the stereoscopic image memory 34 is set to "1".

Next, in step 4, the 2-vertical three-dimensional image obtained in step 2 is transformed by the affine transformation unit 31 as shown in FIG.
Rotate by angle θ in the (x, y) plane, roughly x =
It is rotated by an angle φ within the plane formed by the y straight line and the two axes. Then, in step 5, the circumscribed rectangular parallelepiped extraction unit 3
2 to find a circumscribed rectangular parallelepiped, and the inside of this circumscribed rectangular parallelepiped as ``
1J, and the outside is marked "○".

Next, in step 6, the circumscribed rectangular parallelepiped obtained in the above manner is again transformed by the affine transformation unit 31 inversely to the step 4, and rotated by angles -φ and 1θ. After that, in step 7, the logical product calculation unit 33
The three-dimensional stereo image logical product with the three-dimensional stereo image stored in the three-dimensional image memory 34 is determined in step 3, and is stored in the storage memory 34 again. Then, in step 8, it is determined whether the angle θ is 90 degrees or less, and if it is 90 degrees or less, a minute angle Δθ is added to the angle θ as shown in step 9, and the process returns to step 4 again. Perform steps 4 to 8.

In this way, if the angle θ is 90 degrees or more in step 8, it is determined in step 10 whether the angle φ is 90 degrees or more. If the angle is 90 degrees or less, as shown in step 11, a minute angle Δ is added to the angle φ.
φ is added, the process returns to step 4, and steps 4 to 10 are performed, and the process is repeated until the angle φ becomes 90 degrees or more.

As a result, in step 12, the resultant stereoscopic image obtained by the above-described processing, that is, the outline stereoscopic image that fills the concave portions of the three-dimensional stereoscopic image, is displayed on the resultant stereoscopic image output display 35.

In this way, according to this embodiment, a three-dimensional binary stereoscopic image (
x, y) plane and φ degree relative to the Z axis to obtain a circumscribed rectangular parallelepiped, and then calculate each of these angles θ (0 to 90 degrees) and φ (0 to 90 degrees). Since the contour of the three-dimensional 3D image is determined by calculating the logical product of the circumscribed rectangular parallelepipeds, the contour of the 3D image can be determined by hardware processing.Moreover, since the method does not track the contour of the 3D image, it is possible to calculate the contour of the 3D image at high speed. You can find the outline of the image.

Therefore, by determining the difference between the three-dimensional image obtained by this device and the original three-dimensional image, the pores of the original three-dimensional image and the volume of the pores 1
It becomes possible to obtain a value indicating the surface area or the extent of the recess, and furthermore, it is possible to classify the three-dimensional image based on the characteristic values of the pores or the recess.

Note that the present invention is not limited to the above embodiments.

For example, in the above embodiment, the case where the circumscribed rectangular parallelepiped is obtained by three-dimensionally rotating the three-dimensional stereoscopic image to be measured is shown, but the circumscribed rectangular parallelepiped can be directly calculated from the angles θ and φ without rotating the target three-dimensional image. You may ask for it. Further, in the above embodiment, the case where the angles θ and φ are rotated from 0 to 90 degrees is shown, but it is also possible to rotate the angles θ and φ from 0 to 180 degrees and re-project the projections from each direction onto the stereoscopic image memory 34. good. It goes without saying that various other modifications can be made without departing from the gist of the present invention.

〔Effect of the invention〕

As described in detail above, according to the present invention, it is possible to provide a boundary extraction device that can extract the outline of a three-dimensional 21J stereoscopic image at high speed using simple image processing hardware.

[Brief explanation of drawings]

1 to 4 are diagrams showing an embodiment of the present invention, in which FIG. 1 is a schematic configuration diagram of the device of the present invention, FIG. 2 is a specific configuration diagram of the device of the present invention, and FIG. 3 and 4 are a flowchart and a schematic diagram for explaining the present invention in detail. 11... Three-dimensional stereoscopic image input means, 12... Three-dimensional stereoscopic image binarization means, 13... Three-dimensional stereoscopic image rotation means, 1
4... Circumscribed rectangular parallelepiped extraction means, 15... Three-dimensional stereo image AND operation means, 21... Control computer, 24
...Main memory section, 25...Auxiliary memory section, 26...3
Dimensional stereoscopic image input device, 27... Three-dimensional stereoscopic image processing control unit, 30... Stereo (112 value conversion unit, 31... Affine transformation unit, 32... Circumscribed rectangular parallelepiped extraction unit, 33... Logical product calculation unit, 34...Stereoscopic image memory, 35...Display for outputting the resultant stereoscopic image. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Figure 2 Figure 3

Claims (2)

[Claims] (1) A 3D stereoscopic image input means for inputting a 3D stereoscopic image, and a 3D stereoscopic image binary value for binarizing the 3D stereoscopic image input by the 3D stereoscopic image input means at a predetermined threshold value. 3D stereoscopic image binarization means and 3D images obtained by this 3D stereoscopic image binarization means.
A three-dimensional three-dimensional image rotation means for rotating a dimensional binary three-dimensional image by a predetermined angle θ with respect to the (x, y) plane and a predetermined angle φ with respect to the z-axis; A circumscribed cuboid extraction means for obtaining a circumscribed cuboid of the three-dimensional binary stereoscopic image, and each angle θ (0 to 90 degrees) and φ (0 to 90 degrees) obtained by this circumscribed cuboid extraction means.
1. A boundary extraction device comprising: logical product operation means for calculating a logical product of circumscribed rectangular parallelepipeds and extracting a boundary of a three-dimensional binary stereoscopic image. (2) Claim (1) characterized in that the three-dimensional stereoscopic image input means inputs a three-dimensional stereoscopic image using an X-ray CT scanner or an ultrasonic tomography device.
Boundary extraction device as described in section.