TWI549095B - Real - time map of three - dimensional image generation method - Google Patents

Description

Instant map 3D image generation method

The invention relates to a method for generating an image, in particular to a method for generating a three-dimensional image of a real-time map.

In general, the content of a solid image can only be rendered in a planar manner, or in a two-dimensional manner; even if a general camera or camera is used to capture the image of the solid image, in the captured image The entity image is still rendered in two dimensions. Therefore, if an image that renders the physical image content in a three-dimensional manner is generated according to the content of the entity image, it is necessary to add fun when viewing the physical image content.

For example, if a picture completed by a child in an art course can be presented in a three-dimensional image, it not only enhances the fun of viewing the picture, but also enhances the children's interest in learning.

Accordingly, it is an object of the present invention to provide a three-dimensional image generating method for instant mapping.

Therefore, the method for generating a three-dimensional image of the instant map of the present invention is applicable to a solid two-dimensional image including at least one target object, and comprises a system including an image capturing unit, a processing unit, a display unit, and a database. carried out. The database includes at least one pair of two-dimensional images that should be physically Refer to the 2D image and a pair of reference 3D object models that should reference the 2D image.

The method for generating a three-dimensional image of the instant map comprises five steps: step (a): capturing, by the image capturing unit, a two-dimensional first target image including the target object.

Step (b): the processing unit selects the reference two-dimensional image corresponding to the solid two-dimensional image by comparing the first target image with each reference two-dimensional image of the database, and according to the The reference two-dimensional object model corresponding to the two-dimensional image is selected with reference to the two-dimensional image.

Step (c): capturing, by the processing unit, a partial image area of the first target image that includes the target object.

Step (d): mapping, by the processing unit, the captured partial image area to the reference three-dimensional object model.

Step (e): The processing unit is configured to superimpose the reference three-dimensional object model with the first target image to generate a three-dimensional first superimposed image that can be displayed on the display unit.

The invention has the effect that the fun of viewing the content of the solid two-dimensional image can be added by generating the three-dimensional first superimposed image according to the solid two-dimensional image.

1‧‧‧ system

11‧‧‧Image capture unit

12‧‧‧Processing unit

13‧‧‧Display unit

14‧‧‧Database

141‧‧‧Reference 2D images

1411‧‧‧Line pattern

2‧‧‧ entity two-dimensional image

21‧‧‧ Target objects

4‧‧‧First target image

41‧‧‧Partial image area

411‧‧‧ angular feature points

5‧‧‧First superimposed image

6‧‧‧Second superimposed image

7‧‧‧Mapped reference 3D object model

Other features and effects of the present invention will be apparent from the following description of the drawings. FIG. 1 is a block diagram illustrating a system for performing a three-dimensional image generating method of the instant map of the present invention; 2A is a schematic diagram illustrating the content of a reference two-dimensional image; FIG. 2B is a schematic diagram illustrating the content of another reference two-dimensional image; FIG. 2C is a schematic diagram illustrating the content of another reference two-dimensional image; FIG. A schematic diagram illustrating a physical two-dimensional image including a target object; FIG. 4 is a flow chart illustrating a flow of steps of the method for generating a three-dimensional image of the instant map; FIG. 5 is a schematic diagram illustrating an image capture of the system a first target image of the solid two-dimensional image captured by the unit; FIG. 6 is a schematic diagram illustrating a plurality of angular feature points located in the first target image; FIG. 7 is a schematic diagram illustrating a three-dimensional image A first superimposed image displayed on a display unit of the system; and FIG. 8 is a schematic view illustrating a three-dimensional second superimposed image displayable on the display unit.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are denoted by the same reference numerals.

Referring to FIG. 1, a three-dimensional image generating method of the instant map of the present invention is executed in the system 1 shown in FIG. 1. The system 1 is, for example, a mobile device, and includes an image capturing unit 11, a processing unit 12, a display unit 13, and a database 14. The database 14 includes a plurality of reference two-dimensional images, and a reference three-dimensional object model and a corner feature data corresponding to each of the reference two-dimensional images. Referring to Figures 2A to 2C, Figures 2A to 2C illustrate three parameters. The two-dimensional image 141 is examined, wherein the content of each of the reference two-dimensional images 141 includes a line pattern 1411 drawn by lines.

The three-dimensional image generating method of the instant map is applicable to a solid two-dimensional image including at least one target object. Referring to FIG. 3, FIG. 3 illustrates that the solid two-dimensional image 2 includes the line pattern 1411 shown in FIG. 2A, and the target object 21 is a result of coloring the line pattern 1411. The following takes the physical two-dimensional image 2 shown in FIG. 3 as an example to describe the steps of the three-dimensional image generating method of the instant map.

Referring to FIG. 4 and FIG. 5, first, in step 31, the image capturing unit 11 captures a two-dimensional first target image 4 including the target object 21.

Then, in step 32, the processing unit 12 selects a reference two-dimensional image corresponding to the solid two-dimensional image 2 by comparing each of the first two-dimensional images of the first target image 4 and the database 14 and selects A reference three-dimensional object model corresponding to the reference two-dimensional image of the solid two-dimensional image 2 is output.

The processing unit 12 first converts the first target image 4 into a grayscale image, and performs corner detection according to the grayscale image to calculate a pair of angular feature data of the first target image 4. . Then, the processing unit 12 calculates, for each reference two-dimensional image, a similarity between the angular feature data corresponding to the first target image 4 and the angular feature data corresponding to the reference two-dimensional image; and then the processing unit 12 selects the similarity The reference two-dimensional image corresponding to one of the largest similarities is the reference two-dimensional image corresponding to the solid two-dimensional image 2. In this example, the reference shown in Figure 2A The two-dimensional image 141 is the selected reference two-dimensional image.

Next, in step 33, the processing unit 12 captures a partial image area of the first target image 4 that includes the target object 21.

Referring to FIG. 6 , the angular feature data includes a plurality of angular feature points 411 located in the first target image 4 , and the partial image region 41 is a rectangular region including the isometric feature points 411 . Wherein, if there are n angular feature points 411, and the coordinate positions of the isometric feature points 411 in the first target image 4 are (X1, Y1), (X2, Y2), ..., and (Xn, Yn, respectively) ). Let Xa1=min(X1,X2,...,Xn), Xa2=max(X1,X2,...,Xn), Ya1=min(Y1,Y2,...,Yn), Ya2=max(Y1,Y2,..., Yn), the rectangular area is a rectangular area in which the line segments formed by (Xa1, Ya1) and (Xa2, Ya2) are diagonal.

Next, in step 34, the processing unit 12 maps the captured partial image area to the selected reference three-dimensional object model.

Next, in step 35, referring to FIG. 7, the processing unit 12 superimposes the mapped reference three-dimensional object model 7 with the first target image to generate a three-dimensional first stack that can be displayed on the display unit 13. Image 5 is combined.

Next, in step 36, the image capturing unit 11 captures a two-dimensional second target image including the target object 21.

Then, in step 37, the processing unit 12 calculates a movement information, a rotation information, and a zooming information related to the target object 21 according to the first target image and the second target image.

Finally, at step 38, together with FIG. 8, the processing unit 12 generates a three-dimensional second superimposed image 6 that can be displayed on the display unit 13 according to the movement information, the rotation information, and the zoom information. That is, the processing unit 12 can track the movement, rotation, and zooming of the target object 21 in the first target image 4, and correspondingly move, rotate, or zoom the image in the second superimposed image 6. Reference 3D object model 7.

In summary, the method for generating a three-dimensional image of the instant map of the present invention can continuously generate a corresponding three-dimensional image by continuously capturing an image of the target object 21 in the solid two-dimensional image 2, and the generated three-dimensional image is The mapped reference three-dimensional object model 7 can correspondingly move with the movement, rotation, or scaling of the target object 21 in the solid two-dimensional image 2, so that the content of the two-dimensional image 2 of the entity can be added. It is fun, so it is indeed possible to achieve the object of the present invention.

However, the above is only the embodiment of the present invention, and the scope of the present invention is not limited thereto, that is, the simple equivalent changes and modifications made by the patent application scope and the patent specification of the present invention are still It is within the scope of the patent of the present invention.

31~38‧‧‧Steps

Claims (4)

A method for generating a three-dimensional image of a real-time map, which is applicable to a solid two-dimensional image including at least one target object, and is executed by a system including an image capturing unit, a processing unit, a display unit, and a database. The database includes a plurality of reference two-dimensional images and a plurality of reference three-dimensional object models respectively corresponding to the reference two-dimensional images, and the three-dimensional image generating method of the instant map comprises the following steps: (a) by using the image capturing unit, Extracting a two-dimensional first target image including the target object; (b) calculating, by the processing unit, a similarity between the first target image and the reference two-dimensional image for each reference two-dimensional image, And selecting a reference two-dimensional image corresponding to one of the similarities as a reference two-dimensional image corresponding to the two-dimensional image of the entity, and selecting a reference two-dimensional image corresponding to the two-dimensional image of the entity a reference three-dimensional object model; (c) by the processing unit, capturing a partial image region of the first target image that includes the target object; (d) by the processing unit, capturing the image Mapping the partial image area to the selected reference three-dimensional object model; (e) by the processing unit, superimposing the mapped reference three-dimensional object model and the first target image to generate a three-dimensional displayable on the display unit The first superimposed image; (f) by the image capturing unit, capturing a two-dimensional second target image including the solid two-dimensional image, wherein the second target image is captured The sequence is after the first target image; (g) calculating, by the processing unit, a movement information, a rotation information, and a zoom related to the target object according to the first target image and the second target image And (h) generating, by the processing unit, a three-dimensional second superimposed image displayable on the display unit according to the movement information, the rotation information, and the zoom information. The method for generating a three-dimensional image of the instant map according to claim 1, wherein the database further comprises a corner feature data corresponding to each reference two-dimensional image, and the step (b) comprises the following sub-steps: (b-1) The processing unit converts the first target image into a grayscale image, and calculates a pair of angular feature data corresponding to the first target image according to the grayscale image; (b-2) by using the processing unit, for each Referring to the two-dimensional image, calculating a similarity between the angular feature data corresponding to the first target image and the angular feature data corresponding to the reference two-dimensional image, and selecting a reference two-dimensional image corresponding to a largest one of the similarities A reference two-dimensional image corresponding to the two-dimensional image of the entity; (b-3) by the processing unit, selecting a reference three-dimensional object model corresponding to the reference two-dimensional image corresponding to the two-dimensional image of the entity. The method for generating a three-dimensional image of the instant map according to claim 2, wherein, in step (b-1), the angular feature data includes a plurality of angular feature points located in the first target image, and in step (c) The partial image area is a rectangular area including the isometric feature points. The method for generating a three-dimensional image of the instant map according to claim 3, wherein, in step (c), if there are n angular feature points, and the coordinate positions of the isometric feature points in the first target image are respectively ( X1, Y1), (X2, Y2), ..., and (Xn, Yn), let Xa1 = min(X1, X2, ..., Xn), Xa2 = max(X1, X2, ..., Xn), Ya1 = min (Y1, Y2, ..., Yn), Ya2 = max(Y1, Y2, ..., Yn), the rectangular area is a rectangle whose diagonal line is formed by (Xa1, Ya1) and (Xa2, Ya2) region.