JPS61127072A - Re-formation system of 3-dimensional picture - Google Patents

Abstract

PURPOSE:To show a more accurate form with the minimum number of contour points by deciding the number of contour points according to the complicatedness of a contour and each cut piece image. CONSTITUTION:A picture is supplied and the contour of the picture is traced by means of a track ball to read the contour coordinates. This process is repeated in the same frequency as the number of sheets of the continuous cut piece images. Then the number of contour points serving as the apexes for production of a surface model in accordance with each cut piece image. Based on this number of contour points, the process to decide the coordinates of the contour points is repeated in the same frequency as the number of the cut piece images. A surface model having the decided contour points as apexes is obtained, and the density is decided on each face of the surface model in response to the surface sloping. Then the see-through conversion is carried out with the surface model. A picture thus re-formed in a 3-dimensional form is displayed to a display device.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to determining contour points for creating a surface model per section image in three-dimensional reconstruction of an object.
In particular, this method is suitable for reconstructing cells, etc., in which the size of the target object differs depending on each section image.

[Background of the invention]

Conventional three-dimensional image reconstruction targets cylindrical objects such as the human torso, so as in the known example, contour points are obtained at equal angles in all section images. The contour points of the image were determined and a three-dimensional image was reconstructed. However, since the size and complexity of the microscope target differs depending on the section image, conventional methods may not be able to express an appropriate shape.

[Purpose of the invention]

An object of the present invention is to provide a three-dimensional image reconstruction method that determines the number of contour points for each section image based on the complexity of the contour, and expresses a more accurate shape with the minimum necessary number. .

[Summary of the invention]

In order to achieve the above object, in the method of the present invention, contour points are determined by introducing an index that increases the number of contour points as the contour becomes more complex or longer. That is, first, the length of one contour is calculated, then an index indicating complexity is calculated, and these two values are used to obtain a new index, and based on this, the number of contour points is determined.

Here, the undefined terms "contour point" and "surface model" will be explained using FIG. 1.

Three-dimensional reconstruction of an object is expressed by pasting together triangular surfaces between each section image, as shown in FIG.

A polyhedron created by such a representation is called a "surface model", and those corresponding to the vertices of the triangle are called "contour points".

[Embodiments of the invention]

An embodiment of the present invention will be described below based on FIG. First, FIG. 2 is an overall configuration diagram of a three-dimensional image reconstruction system to which the present invention is applied. 1 is a processing device that controls image input/output and performs image processing.

2 is a display device that outputs three-dimensional reconstructed images that are input images and output images. 3 is a trackball for moving a superimposed cursor (marker) on the displayed image to read contour coordinates, and 4 is a keyboard for performing various operations. 5 is a disk that stores one contour data, original image data, and reconstructed three-dimensional image.

Next, in the above configuration, the entire flow of three-dimensional image reconstruction is shown in FIG. will be explained using FIG.

First, input the 0 image (processing ioo, hereinafter abbreviated as r processing J) whose overall flowchart is shown in Figure 4, trace the outline using a trackball, and read its coordinates (1
01), this process is repeated for the number of consecutive section images (102), and then.

The number of contour points serving as vertices for creating a surface model is determined according to each section image (103). This will be explained in detail using FIG. 3. Based on this number of contour points,
Determine the coordinates of the contour points (104). Repeat steps 103 and 104 for the number of continuous section images (105). Create a surface model with the determined contour points as vertices, and Add density to the surface and perform perspective transformation (
106), the resulting three-dimensionally reconstructed image is displayed on a display device.

Next, 103 will be explained using FIG. 4.

Since the size and complexity of the object differ depending on each section image, we want to express the contour with the minimum necessary contour points according to these, so the index regarding the length and complexity of the shape KEIJO = L / (4 *XS/L") was introduced. Here, L represents length and S represents area.

Here, the denominator (4πXS/L") is 1 in the case of a circle.
, which is the largest and has the property of becoming smaller as the object becomes more complex, so if you divide the length by this, it will become larger if the length is the same.

Therefore, this index is obtained for each section image, and thereby the number of contour points for each section image is determined.

Here, this effect will be explained using a schematic diagram.

FIGS. 5(a) and 5(b) show the number of contour points assigned to each section image using the one-line method and the conventional method for upper and lower continuous section images. That is, in FIG. 5-(a), an index is calculated for each section image using this method, and based on it,
The upper figure has 12 points, and the lower figure has 4 points, but the fifth
In Fig. 2(b), 8 points are assigned to each section image using the conventional method. In this example, the total number of contour points is the same in both FIGS. 5-(a) and (b), but in this method, the size of each section image is the same.

The number of contour points can be assigned depending on the complexity.

C5! Effect of Brightness] As described above with reference to the embodiments, according to the present invention, a more accurate shape can be expressed with a number of contour points corresponding to the size and complexity of the shape of the object in each section image.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram of a surface model, Fig. 2 is a hardware configuration diagram for three-dimensional image reconstruction, Fig. 3 is a flowchart of a three-dimensional image reconstruction program, and Fig. 4 is a flowchart for explaining the present invention in detail. , FIGS. 5(a) and 5(b) are schematic diagrams of an object for which the number of contour points is determined by the conventional method. 1...Image processing device, 2...Display*position, 3...Track thatch 1 Eye grass 2 Hard #3 logi 4-9 Thatch, 5'' hard

Claims (1)

[Claims] 1. When performing three-dimensional reconstruction of an object using multiple section images and displaying its shape using a surface model, the number of contour points that are the vertices of the surface model per one section image is calculated as follows: A three-dimensional image reconstruction method is characterized in that a determination is made according to each section image using an index representing complexity and the length of an object contour.