Image conversion method, texture mapping method, image conversion device, server-client system, and image conversion program

1. An image conversion method for performing a predetermined image conversion, which is one of image upsizing and image downsizing, on a first image to generate a second image, comprising:

a first step of distinguishing a shadow portion in the first image;

a second step of performing a first conversion process on the shadow portion using a brightness value of a pixel according to details of the predetermined image conversion; and

a third step of performing a second conversion process on a portion other than the shadow portion using a surface normal parameter which represents a surface normal of a photograph subject according to the details of the predetermined image conversion,

wherein the first step includes the steps of:

calculating a surface normal parameter of the first image;

correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion,

wherein results of the processes performed at the second and third steps are used to generate the second image.

2. The image conversion method of claim 1 , further comprising a preprocessing step of processing the first image such that the first image is formed only by a diffuse reflection component.

3. The image conversion method of claim 1 , wherein:

the predetermined image conversion is image upsizing;

the first conversion process is a process of interpolating a brightness value to upsize an image; and

the second conversion process is a process of interpolating a surface normal parameter to upsize an image.

4. The image conversion method of claim 1 , wherein:

the predetermined image conversion is image downsizing:

the first conversion process is a process of subsampling a brightness value to downsize an image; and

the second conversion process is a process of subsampling a surface normal parameter to downsize an image.

5. The image conversion method of claim 4 further comprising, before the third step, the step of correcting a surface normal parameter of the first image to smoothly change over spatial positions.

6. The image conversion method of claim 1 , wherein the surface normal parameter is an angle between a surface normal and a light source direction.

7. A texture mapping method, comprising:

a preprocessing step of placing a texture image on a 3-dimensional CG model object;

a first step of distinguishing a shadow portion in the texture image placed on the object;

a second step of performing a first conversion process on the shadow portion using a brightness value of a pixel according to details of a predetermined image conversion, which is one of image upsizing and image downsizing; and

a third step of performing a second conversion process on a portion other than the shadow portion using a surface normal parameter which represents a surface normal of a photograph subject according to the details of the predetermined image conversion,

wherein the first step includes the steps of:

calculating a surface normal parameter of the texture image;

correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion,

wherein results of the processes performed at the second and third steps are used to generate an image of the object.

8. An image conversion device for performing a predetermined image conversion, which is one of image upsizing and image downsizing, on a first image to generate a second image, comprising:

a shadow distinguishing section for distinguishing a shadow portion in the first image;

a first conversion section for performing a first conversion process on the shadow portion distinguished in the shadow distinguishing section using a brightness value of a pixel according to details of the predetermined image conversion; and

a second conversion section for performing a second conversion process on a portion other than the shadow portion using a surface normal parameter which represents a surface normal of a photograph subject according to the details of the predetermined image conversion,

wherein the shadow distinguishing section includes:

means for calculating a surface normal parameter of the first image;

means for correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

means for calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

means for calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion,

wherein results of the processes performed by the first and second conversion sections are used to generate the second image.

9. The image conversion device of claim 8 , wherein the device is implemented as a server-client system

in which the shadow distinguishing section is included in a server; and

the first and second conversion sections are included in a client.

10. An image conversion program stored in a computer readable medium, wherein the image conversion program instructs a computer to perform a method for performing a predetermined image conversion, which is one of image upsizing and image downsizing, on a first image to generate a second image, the method comprising:

a first step of distinguishing a shadow portion in the first image;

a second step of performing a first conversion process on the shadow portion using a brightness value of a pixel according to details of the predetermined image conversion;

a third step of performing a second conversion process on a portion other than the shadow portion using a surface normal parameter which represents a surface normal of a photograph subject according to the details of the predetermined image conversion; and

a fourth step of generating the second image using results of the processes performed at the second and third steps,

wherein the first step includes the steps of:

calculating a surface normal parameter of the first image;

correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion.

11. A shadow recognizing method for distinguishing a shadow portion in an image, comprising the steps of:

calculating a surface normal parameter of the image;

correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion.

12. A shadow recognizing device for distinguishing a shadow portion in an image, comprising:

means for calculating a surface normal parameter of the image:

means for correcting the calculated surface normal parameter on the assumption that a photograph subject has a smooth surface;

means for calculating a brightness value from the corrected surface normal parameter according to a predetermined illumination equation; and

means for calculating, for each pixel, a difference between the calculated brightness value and an actual brightness value, and recognizing a pixel in which the difference is greater than a predetermined value as being the shadow portion.