# US 2004/0001013 A1 · JPEG Huffman table decoder and method thereof based on binary search technique

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2004/0001013 A1
- Document ID: US20040001013A1
- Status: Published application
- Kind: A1
- Application number: 10/406,489
- Filed: 2003-04-04
- Published: 2004-01-01
- Earliest claimed: 2002-06-26
- Assignees: Samsung Electronics Co., Ltd.
- Inventors: Tae-Sun Kim; Tae-Hwan Park
- Web page: https://patentgrep.com/patent/US20040001013A1

## CPC

- H03M 7/40: Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code
- H03M: CODING; DECODING; CODE CONVERSION IN GENERAL (using fluidic means F15C4/00; optical analogue/digital converters G02F7/00; coding, decoding or code conversion, specially adapted for particular applications, see the relevant subclasses, e.g. G01D, G01R, G06F, G06T, G09G, G10L, G11B, G11C, H04B, H04L, H04M, H04N; ciphering or deciphering for cryptography or other purposes involving the need for secrecy G09C)

## Abstract

A joint photographic expert group (JPEG) Huffman table decoder and a method thereof based on a binary search technique. The JPEG Huffman table decoder performs symbol matching based on the binary search and calculates an address of the matched symbol in code book data. Comparison time and an amount of memory required for searching the symbols are reduced.

## Claims

### Claim 1

A joint photographic expert group (JPEG) Huffman table decoder comprising:
a code book memory for storing code book data;
a pre-processing unit for receiving JPEG code streams;
a symbol matching unit for performing symbol matching based on a binary search and calculating an address of the symbol in the code book memory; and
a post-processing unit for calculating and outputting a level of the symbol.

### Claim 2 (depends on 1)

The JPEG Huffman table decoder of claim 1, wherein the symbol matching unit includes:
a first register for storing reference codes to be compared with the JPEG code streams;
a second register for storing address information for the reference codes stored in the first register;
a comparator for comparing the JPEG code streams and the reference codes based on the binary search and determining the number of bits of the symbol in response to the comparison result;
a shifter for performing barrel shifting on the symbol; and
an adder for calculating an address of the symbol in the code book data by using the output data of the shifter and the location information of the reference codes.

### Claim 3 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the reference codes are codes having maximum values among the codes having the same bits as the 1-bit through 16-bit symbols defined in a Huffman table.

### Claim 4 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the reference codes are codes having minimum values among the codes having the same bits as the 1-bit through 16-bit symbols defined in a Huffman table.

### Claim 5 (depends on 1)

The JPEG Huffman table decoder of claim 1, wherein the JPEG code streams input from the pre-processing unit are formed in a left-adjusted type.

### Claim 6 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the shifter converts left-adjusted symbols output from the comparator into right-adjusted symbols.

### Claim 7 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the comparator is a one's complement adder.

### Claim 8 (depends on 7)

The JPEG Huffman table decoder of claim 7, wherein the adder is a one's complement adder to complement differences of ones generated by the comparator.

### Claim 9 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the adder calculates the address of the symbol in the code book by calculating a relative location of the reference code and a target symbol and subtracting the relative location from the absolute address of the reference code.

### Claim 10 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the symbol matching unit further includes a multiplexer for outputting pre-code data, formed by substituting N (where N>0) zeros for the left N bits of data of the JPEG code streams input from the pre-processing unit when the left N bits of data is formed of all ones.

### Claim 11 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the first register is separated into at least two sub-registers to increase a decoding rate.

### Claim 12 (depends on 2)

The JPEG Huffman table decoder of claim 2, wherein the symbol matching unit includes at least two comparators to increase a decoding rate.

### Claim 13

A JPEG Huffman table decoding method comprising:
(a) receiving JPEG code streams;
(b) performing symbol matching based on a binary search to determine the number of bits of the symbol;
(c) performing barrel shifting according to the symbol matching result;
(d) generating an address in a code book memory storing a selected symbol; and
(e) calculating and outputting the level of the symbol.

### Claim 14 (depends on 13)

The JPEG Huffman table decoding method of claim 13, wherein the address is calculated by calculating a relative location of a target symbol and a reference code and subtracting the relative location from the absolute address of the reference code in (d).

### Claim 15 (depends on 14)

The JPEG Huffman table decoding method of claim 14, wherein the reference codes are codes having maximum values among the codes having the same bits as the 1-bit through 16-bit symbols defined in a Huffman table.

### Claim 16 (depends on 14)

The JPEG Huffman table decoding method of claim 14, wherein the reference codes are codes having minimum values among the codes having the same bits as the 1-bit through 16-bit symbols defined in a Huffman table.

### Claim 17 (depends on 13)

A JPEG Huffman table decoder performing the method of claim 13.
