Method of encoding and decoding data in blocks, and encoding and

Details Method of encoding and decoding data in blocks, and encoding and Method of encoding and decoding data in blocks, and encoding and

1990-02-09

1992-08-18

Chin, Stephen

Pulse or digital communications

Spread spectrum

Direct sequence

358133, H04L 166

Patent

active

051406217

DESCRIPTION:

BRIEF SUMMARY
BACKGROUND OF THE INVENTION

1. Field of the Invention
The invention relates to a method of encoding and decoding data in blocks, each block being constituted by a fixed number of data words each comprising a maximum of n bits; and relates to an encoding device and a decoding device implementing this method.
2. Description of the Related Art
The blocks can for example be tables of 8.times.8 or 16.times.16 cosine transformation coefficients resulting from the encoding of the series of images by the cosine transformation. Such a table comprises numerous zero coefficients. The non-zero coefficients have probabilities of appearance which decrease as a function of the absolute value of the coefficients. Conventionally, these coefficients are encoded using a variable length encoding, such as a Huffman encoding which consists in representing by shorter code words the coefficients having the highest probabilities of appearance. A Huffman encoding is efficient if there is a great difference between the probabilities of appearance of the most frequent values and those of the least frequent values. On the other hand, this type of encoding is not efficient for encoding words which have probabilities of appearance which are not very different.
Another type of variable length encoding is described in Picture Coding Symposium PCS86, Tokyo, Japan 2-4 Apr. 1986, pages 148-149; and consists in:
reorganizing the series of data words constituting each block by classifying the words in n classes No i, for i=1 to n, the class No i grouping all of the words having i significant bits;
transmitting the words of non-zero value, of each class No i for i=1 to n successively, omitting to transmit the most significant bit, and each transmitted word being accompanied by an address marking the position of the word in its block. This address is a relative address computed by subtracting from the absolute address of the word to be encoded the absolute address of the last word having been previously encoded in the same class as the word to be encoded. Then this relative address is encoded by a Huffmann encoding.
The decoding successively consists, for each class No i, for i=1 to n, and for each transmitted data word, in:
reconstituting the most significant bit;
deriving the absolute address of each transmitted data word from its encoded address;
and consists in reconstituting the block of data words from the transmitted words thus reconstituted, and from their absolute addresses, by inserting words of zero value at the absolute addresses for which no word has been transmitted.


SUMMARY OF THE INVENTION

The object of the invention is to propose an encoding method perfecting this known method, in order to further reduce the amount of transmitted data. The object of the invention is a method consisting in encoding the absolute addresses marking the position of the data words in a block by computing a relative address, as in the known method, and then in reducing the value of this relative address by subtracting from it a number which is less than the value of the relative address and which can be determined, at the moment of decoding, without having to transmit it.
An advantage of this method is a reduction in the quantity of data to be transmitted by reducing the value of the relative addresses; and that it can be implemented by means of a simple device which is able to operate in real time for conventional television images. The average length of the code words obtained by this method is close to the theoretical minimum number in the case in which the data words to be transmitted have a dispersed statistical distribution.
According to the invention, a method for encoding and decoding blocks of data words, each block having a fixed number of data words with a maximum of n bits, a position of a data word within a block being marked diagonally by a series of absolute addresses having consecutive values. The encoding data words for transmission comprising the steps of: classifying data words into n classes, an ith class, where i=1 to n, groups

REFERENCES:
patent: 4365273 (1982-12-01), Yamada et al.
patent: 4694336 (1987-09-01), Kessen et al.
patent: 4821119 (1989-04-01), Gharavi
patent: 4833535 (1989-05-01), Ozeki et al.
patent: 4873573 (1989-10-01), Thomas et al.

Affiliated with

Also associated with

Rating

Method of encoding and decoding data in blocks, and encoding and does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method of encoding and decoding data in blocks, and encoding and, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of encoding and decoding data in blocks, and encoding and will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-1254760