Band compression device for shaded image

• US 4,454,546 A
• Filed: 03/10/1981
• Issued: 06/12/1984
• Est. Priority Date: 03/13/1980
• Status: Expired due to Term

• Current Assignee: Fuji Photo Film Co Limited (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Photo Film Co Limited (Fujifilm Holdings Corporation)
• Inventors: Mori, Sumio
• Primary Examiner(s): Gruber, Felix D.
• Application Number:
  US06/242,209
• Time in Patent Office: 1,190 Days
• Field of Search: 358/133, 358/260, 358/261, 382/56
• US Class Current: 382/238
• CPC Class Codes: H04N 1/41 Bandwidth or redundancy reduction by scanning H04N1/17 ; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00
• CPC Class Codes: H04N 19/93 Run-length coding

Abstract

A band compression method for a shaded image in which a sample input picture element value and an estimation error obtained from an estimation value of the input picture element are converted into quantization values and coded. First codes are temporarily stored in at least one block of buffer memories according to a quantization step number. The estimation errors are converted into quantization values corresponding to the input picture element values. Each quantization value is converted into the first code when the estimation error is quantized corresponding to the estimation error and quantization value. The next picture element value is estimated in accordance with the quantization value.

Claims

1. A band compression method for a shaded image in which a sampled input picture element value and an estimation error obtained from an estimation value of said input picture element value are converted into quantization values and coded, comprising the steps of:

(a) determining a difference between an input picture element value and a present estimate of said input picture element value;

(b) converting said estimation error into a quantization value in correspondence to said input picture element;

(c) converting said quantization value into a quantization code corresponding to said estimation error and quantization value;

(d) estimating the next input picture element value with said quantization value;

(e) storing said quantization code in a buffer memory;

(f) when a block of a predetermined number of said quantization codes has been stored in said buffer memory, reading out said quantization codes from said buffer memory in bit planes, each of said bit planes consisting of like-ordered bits of said quantization codes;

(g) converting the data contained in each of said bit planes into run-length codes by run-length encoding said data contained in each of said bit planes;

(h) transmitting said run-length codes to a receiving station; and

(i) converting receiving run-length codes to a visual image at said receiving station.

2. The method as claimed in claim 1 in which said step (c) of converting said quantization value into said quantization code comprises providing for said quantization code a first bit indicative of the polarity of a respective estimation error and other bits relating only to a quantization step number irrespective of the positive and negative values of said estimation error.

3. The method as claimed in claim 2 in which, in assigning said other bits to the bit positions corresponding only to said quantization step numbers, a most significant one of said other bits is not assigned to quantization step numbers corresponding to quantization values highest in generation frequency among the quantization values of the positive and negative values of said estimation errors.

4. The method as claimed in claim 1 wherein said step (g) of run-length encoding comprises run-length encoding runs of "0"s and "1"s with different codes.

5. The method as claimed in claim 1 or 4 wherein said step (g) of run-length encoding comprises providing a skip code for runs in which all bits are in the same state.

6. A band compression method for a shaded image, comprising the steps of:

repetitively obtaining an estimation error from a sampled input picture element value and forming from a estimation value thereof a quantization code having a predetermined integer first number of bits;

temporarily storing said quantization codes in at least one block of buffer memories having a number of bit planes equal to said number of bits in each of said quantization codes, wherein like-ordered bits of said quantization code stored in said block of memories are stored in same ones of said bit planes;

when a predetermined second number of said quantization codes has been stored in the bit planes of said buffer memories, successively reading out said quantization codes from said buffer memories in an order of said bit planes;

converting data of each of said bit planes into run-length codes;

transmitting said run-length codes to a receiving station; and

converting received run-length codes to a visual image at said receiving station.