background image
ISO/IEC 10918-1 : 1993(E)
The Interpret markers procedure shall decode the markers which may precede the SOF marker, continuing this decoding
until either a SOF or EOI marker is found. If the DHP marker is encountered before the first frame, a flag is set which
selects the hierarchical decoder at the "hierarchical?" decision point. In addition to the DHP marker (which shall precede
any SOF) and the EXP marker (which shall precede any differential SOF requiring resolution changes in the reference
components), any other markers which may precede a SOF shall be interpreted to the extent required for decoding of the
compressed image data.
If a differential SOF marker is found, the differential frame path is followed. If the EXP was encountered in the Interpret
markers procedure, the reference components for the frame shall be upsampled as required by the parameters in the EXP
segment. The upsampling procedure described in J.1.1.2 shall be followed.
The Decode_differential_frame procedure generates a set of differential components. These differential components shall
be added, modulo 2
16
, to the upsampled reference components in the Reconstruct_components procedure. This creates a
new set of reference components which shall be used when required in subsequent frames of the hierarchical process.
J.2.2
Control procedure for decoding a differential frame
The control procedures in Annex E for frames, scans, restart intervals, and MCU also apply to the decoding of differential
frames and the scans, restart intervals, and MCU from which the differential frame is constructed. The differential frame
differs from the frames of Annexes F, G, and H only at the decoder coding model level.
J.2.3
Decoder coding models for differential frames
The decoding models described in Annexes F, G, and H are modified to allow them to be used for decoding of two's
complement differential components.
J.2.3.1
Modifications to the differential frame decoder DCT coding model
Two modifications are made to the decoder DCT coding models to allow them to code differential frames. First, the IDCT
of the differential output is calculated without the level shift. Second, the DC coefficient of the DCT is decoded directly
without prediction.
J.2.3.2
Modifications to the differential frame decoder lossless coding model
One modification is made to the lossless decoder coding model. The difference is decoded directly without prediction. If
the point transformation parameter in the scan header is not zero, the point transform, defined in Annex A, shall be
applied to the differential output.
J.2.4
Modifications to the entropy decoders for differential frames
The decoding of two's complement differences requires one extra bit of precision in the Huffman code table. This is
described in J.1.4. The arithmetic coding models are already defined for the precision needed in differential frames.
142
CCITT Rec. T.81 (1992 E)
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186]