Image encoding method and apparatus

• US 6,937,773 B1
• Filed: 10/06/2000
• Issued: 08/30/2005
• Est. Priority Date: 10/20/1999
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Nozawa, Shingo, Kondo, Toshiaki
• Primary Examiner(s): Chen, Wenpeng
• Application Number:
  US09/685,008
• Time in Patent Office: 1,789 Days
• Field of Search: 382/245, 382/236, 382/240, 382/239, 382/232, 382/248, 382/107, 382/103, 375/240.15, 375/240.23, 375/240.27, 375/240.08, 375/240.11, 375/240.12, 375/240.18, 714/756, 714/822, 714/824
• US Class Current: 382/243
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/139 Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability
• CPC Class Codes: H04N 19/17 the unit being an image region, e.g. an object
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

An image signal is input from an image input unit and is divided into different spatial frequency bands by applying a discrete wavelet transform thereto using a discrete wavelet transformation unit. On the basis of values of spatial frequency components, a region-of-interest extracts a region of interest by obtaining a distribution of motion vectors in the input image. A quantization unit applies quantization processing to the extracted region of interest and different quantization processing to other regions, and an encoder encodes the quantized image signal. Alternatively, motion of an image contained in the input image may be detected and the region of interest may be obtained based upon motion of this image.

Claims

1. An image encoding apparatus for encoding image signals of a plurality of frames, comprising:

input means for inputting an image signal including pixel values of a frame;

transformation means for applying a discrete wavelet transform to the image signal of each frame and outputting transformed coefficients for each frame;

motion detection means for detecting motion of an image based upon the image signals of plural frames;

counting means for counting a number of pixels based upon information indicating motion of the image detected by said motion detection means;

selection means for selecting a method of designating an area of the image based upon information indicating motion of the image detected by said motion detection means, the selection being based upon the number of pixels counted by said counting means, and for designating a region of the image of the frame based upon the information;

quantization means for quantizing the transformed coefficients of each frame so as to differentiate an image quality of the image of the designated region from an image of other regions, and outputting a quantized image signal; and

encoding means for encoding the quantized image signal quantized by said quantization means.

2. The apparatus according to claim 1, wherein said motion detection means detects motion of the image in accordance with a difference between pixel values of two mutually adjacent pixels vertically of the image signal.

3. The apparatus according to claim 1, wherein said motion detection means detects motion of the image in accordance with a difference between pixel values of corresponding pixels in two successive frames of the image signal.

4. The apparatus according to claim 1, wherein said motion detection means includes:

block calculation means for forming the image signal into blocks and calculating motion vectors on a block-by-block basis; and

detection means for detecting motion of the image based upon whether magnitude of a motion vector calculated by said block calculation means is greater than a predetermined value.

5. The apparatus according to claim 1, wherein said quantization means performs quantization upon raising quantization precision of the image region designated by said region designation means.

6. The apparatus according to claim 1, wherein said encoding means decomposes a data sequence, which is supplied from said quantization means, into bit planes, applies binary arithmetic encoding on a per-bit-plane basis and outputs code sequences giving priority to code sequences that correspond to bit planes of higher order bits.

7. An image encoding method for intra-frame encoding image signals of a plurality of frames, comprising:

an input step of inputting an image signal including pixel values of a frame;

a transformation step of applying a discrete wavelet transform to the image signal of each frame and outputting transformed coefficients of the each frame;

a motion detection step of detecting motion of an image based upon the image signals of plural frames;

a counting step of counting a number of pixels based upon information indicating motion of the image detected in said motion detection step;

a selection step of selecting a method of designating an area of the image based upon the information indicating motion of the image detected in said motion detection step, the selection being based upon the number of pixels counted in said counting step, and for designating a region of the image of the frame based upon the information;

a quantization step of quantizing the transformed coefficients of each frame so as to differentiate an image quality of an image of the designated region from an image of other region, and outputting a quantized image signal; and

and encoding step of encoding the quantized image signal quantized in said quantization step.

8. The method according to claim 7, wherein said motion detection step detects motion of the image in accordance with a difference between pixel values of two mutually adjacent pixels vertically of the image signal.

9. The method according to claim 7, wherein said motion detection step detects motion of the image in accordance with a difference between pixel values of corresponding pixels in two successive frames of the image signal.

10. The method according to claim 7, wherein said motion detection step includes:

a block calculation step of forming the image signal into blocks and calculating motion vectors on a block-by-block basis; and

a detection step of detecting motion of the image based upon whether magnitude of a motion vector calculated at said block calculation step is greater than a predetermined value.

11. The method according to claim 7, wherein said quantization step performs quantization upon raising quantization precision of the image region designated at said region designation step.

12. The method according to claim 7, wherein said encoding step decomposes a data sequence, which is supplied by said quantization step, into bit planes, applies binary arithmetic encoding on a per-bit-plane basis and outputs code sequences giving priority to code sequences that correspond to bit planes of higher order bits.

13. A computer-readable storage medium storing a program for implementing an image encoding method of encoding image signals of a plurality of frames, the program comprising the steps of:

inputting an image signal including pixel values of a frame;

applying a discrete wavelet transform to the image signal of each frame and outputting transformed coefficients of the each frame;

detecting motion of an image based upon the image signals of plural frames;

counting a number of pixels based upon information indicating motion of the image detected in said detecting step;

selecting a method of designating an area of the image based upon information indicating motion of the image detected in said detecting step, the selection being based upon the number of pixels counted in said counting step, and of designating a region of the image of the frame based upon the information;

quantizing the transformed coefficients of each frame so as to differentiate an image quality of the image of the designated region from an image of other regions, and outputting a quantized image signal; and

encoding the quantized image signal quantized in said quantizing step.