Encoder including an error decision circuit

• US 5,274,466 A
• Filed: 01/03/1992
• Issued: 12/28/1993
• Est. Priority Date: 01/07/1991
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Ida, Takashi, Dachiku, Kenshi
• Primary Examiner(s): BRINICH, STEPHEN M
• Application Number:
  US07/816,651
• Time in Patent Office: 725 Days
• Field of Search: 358/133, 358/136, 358/426, 358/433, 382/43, 382/56
• US Class Current: 358/426.09
• CPC Class Codes: H04N 19/00 Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
• CPC Class Codes: H04N 19/98 Adaptive-dynamic-range coding [ADRC]
• CPC Class Codes: H04N 19/99 involving fractal coding

Abstract

An encoder comprises a memory for storing a to-be-coded original image, outputting a to-be-coded image signal in response to a to-be-coded region designation signal for designating a to-be-coded region used as a to-be-coded block, and outputting a to-be-transformed image signal in response to a to-be-transformed region designation signal for designating a to-be-transformed region; a transforming circuit for subjecting the to-be-transformed image signal to a predetermined transformation in response to a transformation method designation signal, and outputting a transformed image signal; a deciding circuit for deciding an error of the transformed image signal with reference to the to-be-coded image signal, and a control circuit for outputting, as a code, information on at least the to-be-transformed region and the transformation method which reduce the error decided by the deciding circuit to a predetermined value or below.

Claims

1. An encoder comprising:

memory means for storing an original image to be coded, said image being divided into a plurality of to-be-coded regions each being a to-be-coded block and to-be-transformed regions, and each region including a number of pixels each having a predetermined pixel value;

control means for outputting a to-be-coded region designation signal for selectively designating the to-be-coded regions for reading out, from said memory means, to-be-coded image data corresponding to one of the to-be-coded regions, outputting a to-be-transformed region designation signal for selectively designating the to-be-transformed regions for reading out, from said memory means, to-be-transformed image data corresponding to one of the to-be-transformed regions, and outputting a transform method designation signal for selectively designating transform methods for transforming the to-be-transformed image data;

transforming means for subjecting the to-be-transformed image data to a predetermined transformation in accordance with the transform method designation signal and outputting the transformed image data;

decision means for deciding an error of the transformed image data with reference to the to-be-coded region image data; and

mean value outputting means for outputting a mean value of pixels of the to-be-coded region as a code,

wherein said control means includes code signal outputting means for outputting, as a code, the information on at least the to-be-transformed region and the transformation method with which the error decided by said decision means is decreased than a predetermined value.

2. The encoder according to claim 1, wherein said transforming means includes means for performing at least one of transformation of the number of pixels, transformation of the pixel value, and transformation of pixel arrangement.

3. The encoder according to claim 1, wherein said transforming means includes means for calculating a mean value of the pixels of each of a plurality of sub-regions obtained by dividing the to-be-transformed region, and processing the to-be-transformed image data for forming the transformed image by means of a plurality of mean values corresponding to said sub-regions.

4. The encoder according to claim 1, wherein said transforming means includes means for processing the to-be-transformed image data, to select a specific pixel from the pixels of each of the sub-regions and form the transformed image data by means of the pixel values of a plurality of specific pixels corresponding to the subregions.

5. The encoder according to claim 1, wherein said transforming means includes means for processing the to-be-transformed image data, to transform the to-be-transformed image data corresponding to the to-be-transformed region having a first shape to the transformed image data corresponding to a region having a second shape differing from the first shape.

6. The encoder according to claim 1, wherein said transforming means includes means for calculating a mean value of the pixel values of each of a plurality of to-be-transformed sub-regions each having the first shape, the to-be-transformed sub-regions being obtained by dividing the to-be-transformed region image having the first shape, and means for forming the transformed image data having the second shape by means of a plurality of mean values corresponding to the to-be-transformed sub-regions.

7. The encoder according to claim 1, wherein said transforming means includes means for processing the to-be-transformed image data, to rotate each of the to-be-transformed regions by a predetermined angle and transform an arrangement of the pixels.

8. The encoder according to claim 1, wherein said control means has a memory for storing a plurality of to-be-transformed region designation data for designating the to-be-transformed regions of the original image stored in said memory means and a plurality of transform method designation data, means for designating said memory means and said transformation means by means of the designation data read out from said memory, to select a combination of a new to-be-transformed region and a new transformation method, and a memory for storing the decision result of the decision means.

9. The encoder according to claim 1, comprising mean value calculating means for calculating a mean value of the pixel values of the transformed image obtained by said transforming means, subtraction means for subtracting the mean value obtained by the mean value calculating means from the transformation image and outputting the subtraction result, and adding means for adding the subtraction result to the mean value and outputting the addition value to the decision means.

10. The encoder according to claim 1, further comprising mean value calculating means for calculating a mean value of the pixel values of the transformed image obtained by said transforming means, first subtraction means for subtracting the mean value from the transformation image and outputting a subtraction result, and second subtraction means for subtracting the mean value from the to-be-coded image data and outputting a second subtraction result, and said decision means decides the error from the first and second subtraction results.

11. The encoder according to claim 10, further comprising deviation ratio calculating means, connected to the first and second subtraction means, for calculating a ratio of a standard deviation of the pixel values of the to-be-coded image data to a standard deviation of the pixel values of the transformed image data from the first and second subtraction results, and multiplying means for multiplying the second subtraction result with the deviation radio and outputting the multiplication result to said decision means, and said decision means decides the error from the first subtraction result and the multiplication result.

12. The encoder according to claim 1, further comprising to-be-coded region/to-be-transformed region address generating means, controlled by said control means, for outputting an address signal for designating the address of said memory means at a predetermined timing, selector means for supplying the address signal of one of said to-be-coded region/to-be-transformed region address generating means, and a subtracter for subtracting the to-be-coded region data read out from said memory means from the transformed image data obtained from said transformation means and sending the subtraction result to said control means.

13. The encoder according to claim 1, wherein said transformation means comprises a memory for storing pixels of the to-be-coded region, and means for reading out the pixels from said memory according to a pattern of the transformation method.