To perform a high-quality image forming process with highly versatile characteristics more efficiently and at higher speed.
A regression coefficient calculation unit 21 learns a coefficient of a regression estimation computation expression for predicting a teacher image pixel value corresponding to a pixel of interest from the pixel of interest of a student image and pixel values around the pixel of interest. A regression estimation unit 23 obtains a tap constituted of the pixel of interest of the student image and the pixel values around the pixel of interest to calculate an estimation value. On the basis of computation results of the regression estimation unit 23, a labelling unit 24 classifies pixels of the student image into a discrimination class A and a discrimination class B. A discrimination coefficient learning unit 25 learns a coefficient to be used for computing the estimation value for determining the discrimination class A and the discrimination class B from the pixel of interest of the student image and the pixel values around the pixel of interest. A discrimination estimation section 27 calculates the estimation value by obtaining the tap constituted of the pixel of interest of the student image and the pixel values around the pixel of interest. On the basis of estimation results of the discrimination estimation unit 27, a classification unit 28 classifies pixels constituting the student image into pixels belonging to the discrimination class A and pixels belonging to the discrimination class B.
SHUDO YASUHIRO
TAKAHASHI NORIAKI
KONDO TETSUJIRO
| JPH0983961A | 1997-03-28 | |||
| JPH08275119A | 1996-10-18 | |||
| JPH0779418A | 1995-03-20 | |||
| JPH0983961A | 1997-03-28 | |||
| JPH08275119A | 1996-10-18 | |||
| JPH0779418A | 1995-03-20 |
Takashi Nishikawa