**ISO/IEC 10918-1 : 1993(E)**

**3.1.6**

**arithmetic decoder:**

An embodiment of arithmetic decoding procedure.

**3.1.7**

**arithmetic encoder:**

An embodiment of arithmetic encoding procedure.

**3.1.8**

**baseline (sequential):**

A particular sequential DCT-based encoding and decoding process specified in this

Specification, and which is required for all DCT-based decoding processes.

**3.1.9**

**binary decision: **

Choice between two alternatives.

**3.1.10**

**bit stream:**

Partially encoded or decoded sequence of bits comprising an entropy-coded segment.

**3.1.11**

**block:**

An 8

×

8 array of samples or an 8

×

8 array of DCT coefficient values of one component.

**3.1.12**

**block-row:**

A sequence of eight contiguous component lines which are partitioned into 8

×

8 blocks.

**3.1.13**

**byte:**

A group of 8 bits.

**3.1.14**

**byte stuffing:**

A procedure in which either the Huffman coder or the arithmetic coder inserts a zero byte into

the entropy-coded segment following the generation of an encoded hexadecimal X'FF' byte.

**3.1.15**

**carry bit:**

A bit in the arithmetic encoder code register which is set if a carry-over in the code register overflows

the eight bits reserved for the output byte.

**3.1.16**

**ceiling function:**

The mathematical procedure in which the greatest integer value of a real number is obtained

by selecting the smallest integer value which is greater than or equal to the real number.

**3.1.17**

**class (of coding process):**

Lossy or lossless coding processes.

**3.1.18**

**code register:**

The arithmetic encoder register containing the least significant bits of the partially completed

entropy-coded segment. Alternatively, the arithmetic decoder register containing the most significant bits of a partially

decoded entropy-coded segment.

**3.1.19**

**coder:**

An embodiment of a coding process.

**3.1.20**

**coding:**

Encoding or decoding.

**3.1.21**

**coding model:**

A procedure used to convert input data into symbols to be coded.

**3.1.22**

**(coding) process:**

A general term for referring to an encoding process, a decoding process, or both.

**3.1.23**

**colour image:**

A continuous-tone image that has more than one component.

**3.1.24**

**columns:**

Samples per line in a component.

**3.1.25**

**component:**

One of the two-dimensional arrays which comprise an image.

**3.1.26**

**compressed data:**

Either compressed image data or table specification data or both.

**3.1.27**

**compressed image data:**

A coded representation of an image, as specified in this Specification.

**3.1.28**

**compression:**

Reduction in the number of bits used to represent source image data.

**3.1.29**

**conditional exchange:**

The interchange of MPS and LPS probability intervals whenever the size of the LPS

interval is greater than the size of the MPS interval (in arithmetic coding).

**3.1.30**

**(conditional) probability estimate:**

The probability value assigned to the LPS by the probability estimation

state machine (in arithmetic coding).

**3.1.31**

**conditioning table:**

The set of parameters which select one of the defined relationships between prior coding

decisions and the conditional probability estimates used in arithmetic coding.

**3.1.32**

**context:**

The set of previously coded binary decisions which is used to create the index to the probability

estimation state machine (in arithmetic coding).

**3.1.33**

**continuous-tone image:**

An image whose components have more than one bit per sample.

**3.1.34**

**data unit:**

An 8

×

8 block of samples of one component in DCT-based processes; a sample in lossless processes.

**2**

**CCITT Rec. T.81 (1992 E)**