Image processing apparatus and method

• US 6,009,213 A
• Filed: 04/23/1997
• Issued: 12/28/1999
• Est. Priority Date: 04/25/1996
• Status: Expired due to Term

• Current Assignee: Canon Ayutthaya Limited (Canon Inc.)
• Original Assignee: Canon Ayutthaya Limited (Canon Inc.)
• Inventors: Miyake, Nobutaka
• Primary Examiner(s): Au, Amelia
• Assistant Examiner(s): Wu, Jingge
• Application Number:
  US08/847,760
• Time in Patent Office: 979 Days
• Field of Search: 382/252, 382/266, 382/270, 382/298, 382/299, 382/300, 358/455, 358/458, 358/466
• US Class Current: 382/300
• CPC Class Codes: G06T 3/403 Edge-driven scaling; Edge-based scaling

Abstract

An image processing apparatus includes a window setting unit which sets a window for input low-resolution information, including a target pixel and pixels surrounding the target pixel. A detection unit detects MAX and MIN values from among the pixels within the window, and a quantization unit quantizes information relating to each pixel, having n values within the window, into a value smaller than n. A LUT (look-up table) determines a distribution ratio of the MAX and MIN values based on quantized values provided by the quantization unit within the window. A transformation unit synthesizes the MAX and MIN values based on the distribution ratio, and a forming unit forms high-resolution output information from the synthesized values. Thus, input low-resolution information is converted into high-resolution output information, at a high speed, to produce a high quality image.

Claims

1. An image processing apparatus for transforming input low-resolution information into high-resolution information by increasing a number of pixels, said apparatus comprising:

window setting means for setting a window including information relating to a low-resolution target pixel and pixels surrounding the target pixel;

detection means for detecting a maximum value A and a minimum value B from among a plurality of pixels within the window;

quantization means for quantizing information relating to each pixel having n values within the window into a value smaller than n;

determination means for determining a distribution ratio a of the values A and B based on quantized values provided by said quantization means within the window;

transformation means for calculating a transformation value F of the target pixel according to a formula, F=∝

A+(1-∝

) B;

interpolation means for calculating an interpolation value based on the transformation value F by said transformation means; and

formation means for forming high-resolution information based on the interpolation value and values A and B,

wherein said determination means includes a memory for outputting the distribution ratio ∝

in accordance with patterns of the quantized values within the window.

3. An apparatus according to claim 1, wherein said quantization means linearly quantizes values between the values A and B.

4. An apparatus according to claim 1, wherein said quantization means makes a number of quantization levels for the target pixel greater than a number of quantization levels for pixels surrounding the target pixel within the window.

2. An image processing method of transforming input low-resolution information into high-resolution information by increasing a number of pixels, said apparatus comprising:

a window setting step of setting a window including information relating to a low-resolution target pixel and pixels surrounding the target pixel;

a detection step of detecting a maximum value A and a minimum value B from among a plurality of pixels within the window;

a quantization step of quantizing information relating to each pixel having n values within the window into a value smaller than n;

a determination step of determining a distribution ratio ∝

of the values A and B based on quantized values provided by said quantization step within the window;

a transformation step of calculating a transformation value F of the target pixel according to a formula, F=∝

A+(1-∝

) B;

an interpolation step of calculating an interpolation value based on the transformation value F by said transformation step; and

a formation step of forming high-resolution information based on the interpolation value and values A and B,

wherein said determination step outputs the distribution ratio ∝

from a memory in accordance with patterns of the quantized values within the window.

5. A method according to claim 2, wherein said quantization step linearly quantizes values between the values A and B.

6. A method according to claim 2, wherein said quantization step makes a number of quantization levels for the target pixel greater than a number of quantization levels for pixels surrounding the target pixel within the window.