Image simulation method

• US 5,594,850 A
• Filed: 01/25/1994
• Issued: 01/14/1997
• Est. Priority Date: 01/29/1993
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Takizawa, Yuri, Kato, Makoto, Noyama, Hideo, Nohmi, Makoto
• Primary Examiner(s): Nguyen, Phu K.
• Application Number:
  US08/186,312
• Time in Patent Office: 1,085 Days
• Field of Search: 395/133, 395/135, 395/155, 395/161, 345/113, 345/114
• US Class Current: 345/632
• CPC Class Codes: G06T 15/405 using Z-buffer

Abstract

Simultaneously and automatically with creating a CG image by computer graphics, (1) the image is segmented to create a mask image, (2) a mask attribute table is created by data transformation, (3) an intrinsic image (object color component, light source color component, ambient color component) is created by data transformation, and the results are saved. The mask image is used to cut a portion from the computer graphics image and the cut portion is combined with a different image, which may be a natural image for example, to obtain a composite image, and at the same time, transparency processing or other such image processing is conducted on the basis of the attribute data. In addition, the color etc. of the object are modified by changing the object color vector or the like.

Claims

1. A composite image simulation method performed by a computer system, comprising the following steps performed by the computer system of:

gathering computer graphics (CG) image data during creation of a CG image with computer graphics based on a shape model;

projecting the shape model on a projection plane for creating a projected image, assigning different mask numbers to different shape model regions of the projected image and at least another region outside the shape model regions, and saving the regions as masks associated with the mask numbers as a mask image, wherein said projecting, assigning, and saving are data transforming conducted automatically in response to said gathering; and

cutting only a desired image from the computer graphics image by using the mask image and the mask numbers to obtain a cut image, and as a part of image simulation combining the cut image and another image that is independent of the computer graphics image to obtain the composite image, wherein said cutting and combining are image simulation.

2. A composite image simulation method according to claim 1, wherein said data transforming further comprises the steps of

classifying the shape model regions and the another region according to transmittance based on surface information of the computer graphics image,

creating the mask image by writing different ones of the mask numbers into the regions, and

forming tables in memory for describing the transmittance as attribute information of the mask image regions;

and

said combining including transparency processing on the basis of the attribute information.

3. A composite image simulation method according to claim 2, wherein

said creating the mask image removes a transparent portion from a first mask;

and

said combining including assembling the first mask, not including the transparent portion, with the another image to obtain a temporary combined image, and then

assembling the transparent portion with the temporary combined image to obtain the composite image having the another image inside an object of the first mask as seen through the transparent portion.

4. A composite image simulation method according to claim 1, wherein said data transforming includes saving a computer graphics image scale as a scalar quantity correlated to an actual length of a part of the shape model;

and

wherein said combining includes automatically changing the relative sizes of the cut image and the another image to match the computer graphics image scale to a scale of the another image.

5. A composite image simulation method according to claim 4, wherein said saving a computer graphics image scale includes saving a perspective grid of actual lengths for one of the mask image and the another image.

6. A composite image simulation method according to claim 5, wherein said combining includes changing the size of the cut image as the cut image is moved within the composite image, with such changing being on the basis of the perspective grid.

7. A composite image simulation method according to claim 1, wherein said data transforming further includes dividing a polyhedral in the computer graphics image into surfaces based on the shape model, creating the masks in accordance with the surfaces, and saving the masks and surface normals as attributes of the masks;

and

said combining including texture mapping the another image to the cut image by geometrical transformation of the another image using the surface normals.

8. A composite image simulation method according to claim 7, said texture mapping including reorienting the another image by geometrical transformation with respect to the surfaces on the basis of the surface normals.

9. A composite simulation method according to claim 1, wherein said data transforming includes the following steps conducted for each pixel of the computer graphics image for producing an intrinsic image:

calculating a light source color vector from color of a light source and specular reflectivity of an object of the computer graphics image;

calculating a light source color component from the light source color vector and angles among the light source, a surface of the object and an eye position;

calculating an object color vector from the light source color and a dispersion reflectivity;

calculating an object color component from the object color vector and from the angle between the light source and the surface of the object;

defining illumination by light other than light directly from the light source as an ambient color component;

creating the intrinsic image from the light source color component, the object color component and the ambient color component; and

wherein said data transforming further includes saving the intrinsic image.

10. A composite image simulation method according to claim 9, wherein said calculating a light source color component uses a light source attenuation factor.

11. A composite image simulation method according to claim 9, wherein said calculating an object color component uses a light source attenuation factor.

12. A composite image simulation method according to claim 9, wherein said image simulation further includes thereafter changing the light source color vector to change the light source color component.

13. A composite image simulation method according to claim 9, wherein said data transforming further includes saving the dispersion reflectivity as a value separate from the color components;

and

said image simulation including changing texture of the object by changing the dispersion reflectivity and recalculating the object color component.

14. A composite image simulation method according to claim 9, wherein said data transforming further includes saving the specular reflectivity as a value separate from the color components;

and thereafter

said image simulation including changing luster of the object by changing the specular reflectivity and recalculating the light source color component.

15. A composite image simulation method according to claim 1,

wherein said data gathering further includes inputting and imaging one data of written character data, pattern outline data and diagram vector data;

wherein said data transforming further includes creating the masks by dividing the CG image into the regions according to properties required for image simulation, saving the masks, creating data tables describing attributes of the masks, and saving the data tables; and

wherein said combining combines the one data and one of the masks with reference to the data tables to obtain the composite image.

16. An image simulation method performed by a computer, said method including the following data transforming steps conducted simultaneously with a creating of a computer graphics image with computer graphics and based on a shape model in the computer, for producing an intrinsic image and being performed for each pixel of the computer graphics image:

calculating a light source color vector from color of a light source and specular reflectivity of an object of the computer graphics image;

calculating a light source color component from the light source color vector and angles among the light source, a surface of the object and an eye position;

calculating an object color vector from the light source color and a dispersion reflectivity of the object;

calculating an object color component from the object color vector and from the angle between the light source and the surface of the object;

defining illumination by light other than light directly from the light source as an ambient color component; and

separately saving each of the light source color component, the object color component and the ambient color component as the intrinsic image.

17. The method of claim 16, further including

image simulating conducted after said data transforming and including subsequently changing color of the object by changing the object color vector.

18. The method of claim 16, wherein said calculating the light source color component uses a light source attenuation factor.

19. The method of claim 16, wherein said calculating the object color component uses a light source attenuation factor.

20. The method of claim 16, said data transforming further including image simulating conducted after said data transforming and including thereafter changing the light source color vector to change the light source color.

21. The method of claim 20, said data transforming further including saving the dispersion reflectivity as a value separate from the color components;

and

said image simulating including changing texture of the object by changing the dispersion reflectivity and recalculating the object color component.

22. The method of claim 20, said data transforming further including saving the specular reflectivity as a value separate of the color components;

and thereafter

said image simulating including changing luster of the object by changing the specular reflectivity and recalculating the light source color component.

23. An image simulation method comprising the steps of:

inputting and imaging one data of written character data, pattern outline data and diagram vector data;

creating masks by dividing an image into regions according to properties required for image simulation;

saving the masks;

creating data tables describing attributes of the masks;

saving the data tables; and

combining the one data and one of the masks with reference to the data tables to obtain a composite image.

24. A composite image simulation method performed by a computer system, comprising the following steps performed by the computer system of:

gathering computer graphics (CG) image data during creation of a CG image with computer graphics based on a three-dimensional shape model;

projecting the three-dimensional shape model on a projection plane for creating a projected image, assigning different mask numbers to different three-dimensional shape model regions of the projected image and at least another region outside the three-dimensional shape model regions, and saving the regions as masks associated with the mask numbers as a mask image, wherein said projecting, assigning, and saving are data transforming conducted automatically in response to said gathering; and

cutting only a desired image from the computer graphics image by using the mask image and the mask numbers to obtain a cut image, and as a part of image simulation combining the cut image and another image that is independent of the computer graphics image to obtain the composite image, wherein said cutting and combining are image simulation.

25. A composite image simulation method according to claim 24, wherein said data transforming further comprises the steps of

classifying the three-dimensional shape model regions and the another region according to transmittance based on surface information of the computer graphics image,

creating the mask image by writing different ones of the mask numbers into the regions, and

forming tables in memory for describing the transmittance as attribute information of the mask image regions;

and

said combining including transparency processing on the basis of the attribute information.

26. A composite image simulation method according to claim 25, wherein

said creating the mask image removes a transparent portion from a first mask;

and

said combining including assembling the first mask, not including the transparent portion, with the another image to obtain a temporary combined image, and then

assembling the transparent portion with the temporary combined image to obtain the composite image having the another image inside an object of the first mask as seen through the transparent portion.

27. A composite image simulation method according to claim 24, wherein said data transforming includes saving a computer graphics image scale as a scalar quantity correlated to an actual length of a part of the three-dimensional shape model;

and

wherein said combining includes automatically changing the relative sizes of the cut image and the another image to match the computer graphics image scale to a scale of the another image.