JPH03238970A - Conversion coding system - Google Patents
Conversion coding systemInfo
- Publication number
- JPH03238970A JPH03238970A JP2034658A JP3465890A JPH03238970A JP H03238970 A JPH03238970 A JP H03238970A JP 2034658 A JP2034658 A JP 2034658A JP 3465890 A JP3465890 A JP 3465890A JP H03238970 A JPH03238970 A JP H03238970A
- Authority
- JP
- Japan
- Prior art keywords
- coefficients
- zero
- dimensional
- coefficient
- quantization
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000006243 chemical reaction Methods 0.000 title abstract description 17
- 238000013139 quantization Methods 0.000 claims abstract description 45
- 230000009466 transformation Effects 0.000 claims description 48
- 238000000034 method Methods 0.000 claims description 9
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 abstract description 14
- 238000004364 calculation method Methods 0.000 abstract description 6
- 239000011159 matrix material Substances 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000000903 blocking effect Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 206010000210 abortion Diseases 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
Landscapes
- Image Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
【発明の詳細な説明】
［産業上の利用分野］
この発明は、画像データを線形変換符号化方式を用いて
帯域圧縮を行うものに関するものである。DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to band compression of image data using a linear transformation coding method.
［従来の技術］
第３図は例えばＷ、 Ｈ，ＣＨＥＮ、 Ｗ、 Ｋ、 Ｐ
ＲＡＴＴ、 ’５ｃｅｎｅ Ａｄａｐｔｉｖｅ Ｃｏｄ
ｅｒ’、 （ＬＥＥＥ Ｔｒａｎｓａｃｔｉｏｎｓ ｏ
ｎ ｃｏｍｍｕｎｉｃａｔｉｏｎｓ、 ｖｏｌ、 Ｃ０
Ｍ−３２，Ｎｏ、 ３．１９８４）に示された従来の変
換符号化方式を示すブロック図であり、図において、（
１）は入力信号をブロック化するブロック化部、（２）
はブロック化された信号を線形変換する２次元線形変換
部、（３）は信号列をプロツク内で並び換えるスキャン
変換部、（４）は量子化部、（５）は有効無効識別部、
（６）は符号化部、（７）は送信バッファ、（８）は符
号化制御部である。[Prior art] Figure 3 shows, for example, W, H, CHEN, W, K, P.
RATT, '5cene Adaptive Cod
er', (LEEE Transactions o
n communications, vol, C0
3. M-32, No. 3.1984);
1) is a blocking unit that blocks the input signal; (2)
(3) is a scan conversion unit that rearranges the signal sequence within the block; (4) is a quantization unit; (5) is a valid/invalid discriminator;
(6) is an encoding unit, (7) is a transmission buffer, and (8) is an encoding control unit.
次に動作について説明する。ディジタル化され、た１フ
レ一ム分の入力画像信号（１０１）に対し、ブロック化
部（１）で水平、垂直方向Ｎ画素（Ｎは自然数で例えば
Ｎ＝４．８．１６）を１まとめにした２次元の画素ブロ
ックに分割する。ブロック化された画像信号（１０２）
に対し、線形変換部（２）では２次元線形変換（例えば
離散コサイン変換などの直交変換）を施し、空間周波数
領域の変換係数ブロック（１０３）を生成する。ここで
例えば８×８画素をブロック化した行列をｆ１変換行列
をＡとすると、２次元離散コサイン変換係数行列Ｆは次
式で与えられる。Next, the operation will be explained. The blocking unit (1) collects N pixels in the horizontal and vertical directions (N is a natural number, for example, N=4.8.16) from the input image signal (101) for one frame that has been digitized. Divide into two-dimensional pixel blocks. Blocked image signal (102)
On the other hand, the linear transformation unit (2) performs two-dimensional linear transformation (for example, orthogonal transformation such as discrete cosine transformation) to generate a transformation coefficient block (103) in the spatial frequency domain. For example, if the f1 transformation matrix is A, which is a matrix obtained by blocking 8×8 pixels, then the two-dimensional discrete cosine transformation coefficient matrix F is given by the following equation.
Ｆ＝ＡｆＡ” ・・・■
ここで、Ａ↑はＡの転置行列であり、変換行列Ａの要素
は次式で表される。F=AfA"...■ Here, A↑ is the transposed matrix of A, and the elements of the transformation matrix A are expressed by the following equation.
Ａ （ｉ、Ｄ ＝ １／２Ｃ（ｉ）ｃｏｓ ［ｒｉ（２
ｊ＋１）／１６ ］ここで、ｉ、ｊ＝０．ｉ−＋
、７であり、である。A (i, D = 1/2C(i)cos [ri(2
j+1)/16] where i, j=0. i-+
, 7, and .
式■から判るように２次元線形変換は画素行列ｆに対し
て行方向と列方向の２回の１次元線形変換演算を施すこ
とにより得られる。As can be seen from equation (2), the two-dimensional linear transformation is obtained by performing two one-dimensional linear transformation operations on the pixel matrix f, in the row direction and column direction.
変換係数行列Ｆの要素Ｆ（ｕ、 ｖ）（ｕ、 ｖ＝ｏ、
１． ・・＋、 ７）の性質を第４図をもとに説明す
る。Ｆ（ｕ、ｖ）の値はブロック化された画像信号（１
０２）に含まれる空間周波数成分がそれぞれどの程度で
あるかを示している。水平方向の周波数はＵの値が大き
くなるにつれて高くなり、垂直方向の周波数はＶの値が
大きくなるにつれて高くなる。すなわちＦ（０，０）の
値はブロック化された画像信号（１０２）の直流成分の
強度に対応し、Ｆ（７，７）の値は水平・垂直方向とも
に高い周波数をもつ交流成分の強度に対応することにな
る。従って、画素の値の変化が少ない背景などの平坦な
画像ブロックに対しては低周波成分のみに非零の有意係
数が現れ、高周波成分はほとんど零係数となる。逆に画
素の変化が激しいエツジ部分などの画像ブロックに対し
ては低周波成分のほか高周波成分にも非零の有意係数が
現れる。Element F(u, v) of transformation coefficient matrix F (u, v=o,
1. ...+, 7) will be explained based on Figure 4. The value of F(u,v) is the blocked image signal (1
02) shows how much the spatial frequency components are included in each. The frequency in the horizontal direction increases as the value of U increases, and the frequency in the vertical direction increases as the value of V increases. In other words, the value of F(0,0) corresponds to the intensity of the DC component of the blocked image signal (102), and the value of F(7,7) corresponds to the intensity of the AC component with high frequency in both the horizontal and vertical directions. It will correspond to Therefore, for a flat image block such as a background with little change in pixel values, non-zero significant coefficients appear only in low frequency components, and almost zero coefficients appear in high frequency components. On the other hand, for image blocks such as edge portions where pixels change rapidly, non-zero significant coefficients appear in high frequency components as well as low frequency components.
次に、スキャン変換部（３）では変換係数ブロック（１
０３）のブロック内で例えば第４図の矢印で示す順序で
変換係数を並び換え、１次元の変換係数列Ｆ（ｎ）（１
０４）を出力する。先の８×８画素ブロックの場合、１
ブロツクに対し８２−６４個の係数が続く係数列（ｎ＝
１〜６４）が出力され、例えば要素Ｆ（０，０）はＦ（
１）に、Ｆ（７，７）はＦ（６４）になる。並び換えは
非零の有意係数が現われやすい低周波成分の変換係数か
ら有意係数が現われにくい高周波成分の変換係数へとジ
グザグに走査することにより有意係数をなるべく前半に
、後半に零係数を長く続かせるために行う。Next, in the scan conversion unit (3), the conversion coefficient block (1
03), the transform coefficients are rearranged in the order shown by the arrows in Fig. 4, for example, to form a one-dimensional transform coefficient sequence F(n)(1
04) is output. For the previous 8x8 pixel block, 1
A coefficient sequence of 82-64 coefficients (n=
1 to 64) are output, for example, element F(0,0) is F(
1), F(7,7) becomes F(64). The rearrangement is done by scanning in a zigzag pattern from the conversion coefficients of low frequency components where non-zero significant coefficients tend to appear to the conversion coefficients of high frequency components where significant coefficients are unlikely to appear, so that the significant coefficients are placed in the first half as much as possible and the zero coefficients are kept in the second half for a long time. I do it to make it work.
次に、量子化部（４）は変換係数列（１０４）を、後で
述べる与えられた量子化ステップサイズ（１１０）で量
子化し、量子化係数列Ｑ（ｎ） （１０５）を出力する
。有効無効識別部（５）では量子化係数列（１，０５）
がすべて零であるかどうかの判定を行う。全ての係数が
零の場合は無効ブロック、１つでも非零の有意係数があ
る場合は有効ブロックとして有効無効情報（１０６）を
符号化部（６）に出力する。符号化部（６）では有効無
効情報（１０６）により有効ブロックと判定された場合
のみ、量子化係数列（１０５）に符号の割り当てを行い
、符号化データ（１０７）として送信バッファ（７）へ
出力する。これに対し、有効無効情報（１０６）により
無効ブロックと判定された場合には、無効ブロックを表
す符号を符号化データ（１０７）として送信バッファ（
７）へ出力する。Next, the quantization unit (4) quantizes the transform coefficient sequence (104) with a given quantization step size (110), which will be described later, and outputs a quantized coefficient sequence Q(n) (105). In the valid/invalid identification unit (5), the quantized coefficient sequence (1,05)
Determine whether all are zero. Validity/invalidity information (106) is output to the encoding unit (6) as an invalid block if all coefficients are zero, and as a valid block if there is even one non-zero significant coefficient. The encoding unit (6) assigns a code to the quantized coefficient sequence (105) only when the block is determined to be a valid block based on the validity/invalidity information (106), and sends it to the transmission buffer (7) as encoded data (107). Output. On the other hand, if the block is determined to be invalid based on the valid/invalid information (106), the code representing the invalid block is stored in the transmission buffer (107) as encoded data (107).
7).
ここで符号の割り当て方法の１例として２次元可変長符
号化について説明する。これは量子化係数列（１０５）
に対して連続する零係数の個数（以下ゼロランと呼ぶ）
とそれに続く非零係数の量子化レベルを組み合わせ、そ
の組み合わせた事象（ゼロラン、量子化レベル）に対し
て１つのハフマン符号を割り当てることによって行われ
る。Here, two-dimensional variable length coding will be described as an example of a code assignment method. This is a quantized coefficient sequence (105)
The number of consecutive zero coefficients for (hereinafter referred to as zero run)
This is done by combining the quantization levels of the following non-zero coefficients and assigning one Huffman code to the combined event (zero run, quantization level).
第５図は１つのブロックの量子化係数列（１０５）を示
すもので、量子化係数Ｑ（１）、Ｑ（４）、Ｑ（９）、
Ｑ（１３） 、Ｑ（２１）は量子化レベルがそれぞれ２
０．１５．５．２．１であり、その他の量子化係数は零
であるので、事象（ゼロラン、量子化レベル）は次のよ
うになる。FIG. 5 shows the quantization coefficient sequence (105) of one block, including quantization coefficients Q(1), Q(4), Q(9),
Q(13) and Q(21) each have a quantization level of 2.
0.15.5.2.1, and other quantization coefficients are zero, so the event (zero run, quantization level) is as follows.
（０，２０）、 （２，１５）、 （４，５）、 （３
，２）、 （７，１）、 ＥＯＢここでＥＯＢは以降に
非零の有意係数がなく、ブロックの終りまで零係数が続
くことを示すマークである。従って、この量子化係数列
の場合ＥＯＢを含めた６つの事象に対して、それぞれに
決められたハフマン符号が割り当てられることになる。(0,20), (2,15), (4,5), (3
, 2), (7, 1), EOB Here, EOB is a mark indicating that there are no non-zero significant coefficients thereafter, and that zero coefficients continue until the end of the block. Therefore, in the case of this quantized coefficient sequence, a predetermined Huffman code is assigned to each of six events including EOB.
次に送信バッファ（７）では変動する情報発生量を平滑
化し、一定レートで伝送路（１０８）へ送出する。符号
化制御部（８）では送信バッファ（７）中のデータ残量
であるバッファ残量（１０９）から量子化ステップサイ
ズ（１１０）を適応的にフィードバック制御し、量子化
部（４）へ出力する。すなわち、バッファ残量（１０９
）が多いときには、これから発生する情報量を少なくす
るために量子化ステップサイズ（１１，０）を大きくし
て変換係数列（１０４，）を粗く量子化する。Next, the transmission buffer (7) smoothes the fluctuating amount of generated information and sends it to the transmission path (108) at a constant rate. The encoding control unit (8) adaptively controls the quantization step size (110) based on the remaining buffer amount (109), which is the remaining amount of data in the transmission buffer (7), and outputs it to the quantization unit (4). do. In other words, the remaining buffer capacity (109
) is large, the quantization step size (11,0) is increased to coarsely quantize the transform coefficient sequence (104,) in order to reduce the amount of information generated.
逆に、バッファ残量（１０９）が少ないときには、これ
から発生する情報量を多くするために量子化ステップサ
イズ（１１０）を小さくして変換係数列（１，０４）を
細かく量子化する。Conversely, when the remaining buffer capacity (109) is small, the quantization step size (110) is reduced to finely quantize the transform coefficient sequence (1,04) in order to increase the amount of information that will be generated.
［発明が解決しようとする課題］
従来の変換符号化方式は以上のように構成されているの
で、有効無効識別・２次元可変符号化の処理を行うため
に全ての量子化係数が必要であり、そのため全ての２次
元変換係数を求めるための２回のｔ次元線形変換演算と
全ての２次元変換係数に対する量子化処理を行わなけれ
ばならなかった。[Problems to be Solved by the Invention] Since the conventional transform encoding method is configured as described above, all quantization coefficients are required to perform valid/invalid identification and two-dimensional variable encoding processing. Therefore, it was necessary to perform two t-dimensional linear transformation operations to obtain all the two-dimensional transformation coefficients and to perform quantization processing on all the two-dimensional transformation coefficients.
この発明は上記のような問題点を解決するためになされ
たもので、変換係数の伝送範囲を変換係数ブロック内の
量子化係数列に応じて決定し、必要な２次元変換係数の
みを順次１つづつ求め量子化すると共に同時に有効無効
識別・２次元可変長符号化を行うための事象を生成し、
処理に要する演算量・処理時間を削減する変換符号化方
式を得ることを目的とする。This invention was made to solve the above-mentioned problems.The transmission range of transform coefficients is determined according to the quantized coefficient sequence in the transform coefficient block, and only the necessary two-dimensional transform coefficients are sequentially transmitted. Generate events for sequentially determining and quantizing and simultaneously performing valid/invalid identification and two-dimensional variable length encoding,
The purpose is to obtain a transform encoding method that reduces the amount of calculations and processing time required for processing.
［課題を解決するための手段］
この発明に係わる変換符号化方式は、入力信号系列に対
して２次元線形変換を行い変換領域で低域から高域へ変
換係数を順次量子化し符号化する変換符号化方式におい
て、ブロック化された入力信号系列に１次元線形変換を
施し１次元変換係数を得る手段と、（次元変換係数にさ
らに直交する１次元線形変換を施し低域から高域へ順次
１つの２次元変換係数を得る手段と、量子化された２次
元変換係数の値のうち連続する零係数の個数を計数する
手段と、量子化された２次元変換係数のうち非零係数と
その非零係数が現れるまでに計数手段により計数された
連続零係数の個数との組を記憶する手段と、符号化情報
発生量を所定の伝送情報量に近付けるために符号化伝送
する連続零係数の個数の閾値を送信バッファのデータ残
量から設定する手段と、計数された連続零係数の計数値
が閾値を越えたとき後続する２次元変換係数を求めるた
めの１次元線形変換及び量子化処理を打ち切り記憶手段
の記憶内容に対して符号を割当てる手段とを備える。[Means for Solving the Problems] A transform encoding method according to the present invention performs a two-dimensional linear transform on an input signal sequence, and sequentially quantizes and encodes transform coefficients from a low frequency band to a high frequency band in a transform domain. In the encoding method, there is a means for obtaining one-dimensional transform coefficients by performing one-dimensional linear transform on a blocked input signal sequence, and a means for obtaining one-dimensional transform coefficients by performing one-dimensional linear transform (further orthogonal to the dimensional transform coefficients, sequentially from low frequency to high frequency). means for counting the number of consecutive zero coefficients among the values of the quantized two-dimensional transform coefficients; and means for counting the number of consecutive zero coefficients among the values of the quantized two-dimensional transform coefficients; means for storing a set of the number of consecutive zero coefficients counted by the counting means until the zero coefficient appears, and the number of consecutive zero coefficients to be encoded and transmitted in order to bring the amount of generated encoded information close to a predetermined amount of transmitted information. means for setting a threshold value from the remaining amount of data in the transmission buffer, and aborts one-dimensional linear transformation and quantization processing for obtaining subsequent two-dimensional transformation coefficients when the counted value of continuous zero coefficients exceeds the threshold value. and means for assigning a code to the stored contents of the storage means.
［作用］
この発明に係わる変換符号化方式は入力信号ブロックに
対して１次元線形変換を行い、さらに直交する１次元線
形変換を施し低域から高域へ順次１つの２次元変換係数
を得、量子化を行い、連続する零係数の個数を計数し非
零係数値とその非零係数が現れるまでに計数された連続
零係数の個数との組を事象として一時記憶しておくと共
に、連続零係数の個数が送信バッファのデータ残量から
設定された閾値を越えたとき後続する２次元変換係数を
求めるための１次元線形変換及び量子化を打ち切り記憶
された事象に対して符号の割当てを行う。[Operation] The transform encoding method according to the present invention performs one-dimensional linear transformation on an input signal block, and further performs orthogonal one-dimensional linear transformation to sequentially obtain one two-dimensional transformation coefficient from the low frequency band to the high frequency band, Perform quantization, count the number of consecutive zero coefficients, temporarily store the set of the non-zero coefficient value and the number of consecutive zero coefficients counted until the non-zero coefficient appears as an event, and When the number of coefficients exceeds a threshold set from the remaining amount of data in the transmission buffer, the one-dimensional linear transformation and quantization for obtaining the subsequent two-dimensional transformation coefficients are discontinued and a code is assigned to the stored event. .
［発明の実施例］ 以下、この発明の一実施例を第１図をもとに説明する。[Embodiments of the invention] An embodiment of the present invention will be described below with reference to FIG.
図において（９）は１次元線形変換部、（１０）は定め
られた順序により２次元変換係数を１つづつ求める１次
元線形変換部、（１１）は連続した零の量子化係数を計
数するゼロカウンタ、（１２）は閾値を設定する閾値設
定部、（上３）は計数値と閾値とを比較し大小の判定を
行う判定部、（１４）は非零の量子化係数値とそのとき
の計数値の組である事象を一時記憶させる事象記憶部、
（１５）は事象に対して符号の割当てを行う符号割当て
部であり、他は第３図と同様である。In the figure, (9) is a one-dimensional linear transformation section, (10) is a one-dimensional linear transformation section that obtains two-dimensional transformation coefficients one by one in a predetermined order, and (11) is a one-dimensional linear transformation section that counts consecutive zero quantization coefficients. Zero counter, (12) is a threshold setting section that sets a threshold, (3) is a judgment section that compares the counted value with the threshold and determines whether it is large or small, (14) is a non-zero quantization coefficient value and its time an event storage unit that temporarily stores an event that is a set of counted values;
(15) is a code assignment unit that assigns a code to an event; the other parts are the same as in FIG.
また、第２図は動作を説明するためのフローチャート図
である。Moreover, FIG. 2 is a flowchart diagram for explaining the operation.
次に第２図と共に動作について説明する。第３図と同様
、ディジタル化された１フレ一ム分の人力画像信号（１
０１）はブロック化部（１）でＮ×Ｎ画素のブロックに
分割する。ブロック化された画像信号（１０２）は１次
元線形変換部（９）で例えば行方向の１次元線形変換演
算か行われＮ×Ｎ個の要素を持つ１次元変換係数ブロッ
ク（１１１）を得る（ステップ１）。ここで、初期設定
としてゼロカウンタ（１１）の計数値（１，１４）を零
に、事象記憶部（１４）の記憶内容をクリアし、Ｎ２個
ある２次元変換係数を第４図に示されるスキャン順序で
スキャンしたときの係数番号ｉを１とする（ステップ２
）。つぎに１次元線形変換部（１０）では１次元変換係
数ブロック（１１（）に対して今度は直交する列方向の
↓次元線形変換演算が行なわれ、係数番号１の２次元変
換係数Ｆ（ｉ） （１↓２）が１つ出力される（ステッ
プ３）。符号化制御部（８）は送信バッファ（７）のバ
ッファ残量（１０９）から量子化ステップサイズ（１１
０）を決定し、量子化部（４）へ出力する。Next, the operation will be explained with reference to FIG. Similar to Figure 3, the digitalized human image signal for one frame (1
01) is divided into blocks of N×N pixels by the blocking unit (1). The blocked image signal (102) is subjected to a one-dimensional linear transformation operation in the row direction, for example, in a one-dimensional linear transformation unit (9) to obtain a one-dimensional transformation coefficient block (111) having N×N elements ( Step 1). Here, as an initial setting, the count value (1, 14) of the zero counter (11) is set to zero, the memory contents of the event storage unit (14) are cleared, and the N2 two-dimensional conversion coefficients are converted to the values shown in FIG. The coefficient number i when scanned in the scan order is set to 1 (Step 2
). Next, in the one-dimensional linear transformation unit (10), a ↓-dimensional linear transformation operation in the orthogonal column direction is performed on the one-dimensional transformation coefficient block (11()), and the two-dimensional transformation coefficient F(i ) (1↓2) is output (step 3).The encoding control unit (8) calculates the quantization step size (11
0) and outputs it to the quantization section (4).
閾値設定部（１２）では同じくバッファ残量（１０９）
から閾値（１１５）を決定し、判定部（１３）へ出力す
る。量子化部（４）では変換係数Ｆ（ｉ）（１１２）を
量子化ステップサイズ（１１０）で量子化し、量子化係
数Ｑ（ｉ） （１１３）を出力する（ステップ４）。つ
ぎに、この量子化係数Ｑ（ｉ）について零であるか非零
であるかを判定しくステップ５） 、Ｑ（ｉ）の値が零
でない場合、事象記憶部（１４）ではゼロカウンタ（１
１）の計数値（１１４）と非零の係数であるＱ（１）の
組を事象として記憶し、ゼロカウンタ（１１）をリセッ
トして零とする（ステップ６）。一方、上記ステップ５
で量子化係数Ｑ（ｉ）が零の場合、ゼロカウンタ（１１
）の計数値（１，１４）に１が加えられ（ステップ７）
、判定部（１３）でその計数値（１１４）と閾値（１１
，５）との大小比較を行い（ステップ８）、判定結果（
１１６）を出力する。その計数値（１１４）が閾値（１
１５）と等しいかまたは大きいときには、出力された判
定結果（１１６）にもとづいて■次元線形変換部（１０
）および量子化部（４）の処理を打ち切る（ステップ９
）。そして、事象記憶部（１４）に記憶されている量子
化係数の零係数の個数と非零係数のレベルとの組を読み
だし符号割り当て部（１５）へ出力する（ステップ（２
）。また、前述した判定部（１３）での計数値（１１４
）と閾値（１１５）の比較において計数値（１１４）が
閾値（１１５）より小さいときまたはステップ６が終了
したときは、係数番号ｉがＮ２となったかを判定しくス
テップ１０）、係数番号ｉがＮ２以下でＱ（ｉ）が最後
の量子化係数でなければ係数番号１に１を加え（ステッ
プ１１）、次の２次元変換係数Ｆ（ｉ）の演算、量子化
を引き続き行う。係数番号ＩがＮ２、すなわちＱ（ｉ）
が最後の量子化係数である場合、現在記憶されている事
象（１１，７）を出力しくステップ１２）、その画素ブ
ロックの処理を終了する。符号割当て部（１５）は出力
された事象（１１７）に対してハフマン符号の割当てを
行いＥＯＢを付加して、符号化データ（１０７）として
送信バッファ（７）へ出力する。これに対し、出力され
る事象（１１７）がない場合は、無効ブロックであるた
め無効ブロックを表す符号を符号化データ（１０７）と
して送信バッファ（７）へ出力する。In the threshold setting section (12), the remaining buffer amount (109)
A threshold value (115) is determined from the threshold value (115) and output to the determination unit (13). The quantization unit (4) quantizes the transform coefficient F(i) (112) with a quantization step size (110) and outputs a quantization coefficient Q(i) (113) (step 4). Next, it is determined whether this quantization coefficient Q(i) is zero or non-zero (step 5). If the value of Q(i) is not zero, the event storage unit (14) stores a zero counter (1).
The set of the count value (114) of 1) and the non-zero coefficient Q(1) is stored as an event, and the zero counter (11) is reset to zero (step 6). On the other hand, step 5 above
If the quantization coefficient Q(i) is zero, the zero counter (11
) is added to the count value (1, 14) (step 7)
, the count value (114) and the threshold value (11
, 5) (step 8), and the judgment result (
116) is output. The count value (114) is the threshold value (1
15), the ■dimensional linear transformation unit (10
) and quantization unit (4) are terminated (step 9
). Then, the set of the number of zero coefficients and the level of non-zero coefficients of the quantization coefficients stored in the event storage unit (14) is read out and output to the code assignment unit (15) (step (2)
). In addition, the count value (114
) and the threshold value (115), when the count value (114) is smaller than the threshold value (115) or when step 6 is completed, it is determined whether the coefficient number i has become N2 (step 10). If Q(i) is not the last quantized coefficient within N2, 1 is added to coefficient number 1 (step 11), and the calculation and quantization of the next two-dimensional transformation coefficient F(i) are continued. Coefficient number I is N2, i.e. Q(i)
If is the last quantized coefficient, the currently stored event (11,7) is output (step 12) and the processing of that pixel block is ended. The code allocation unit (15) allocates a Huffman code to the output event (117), adds an EOB, and outputs it to the transmission buffer (7) as encoded data (107). On the other hand, if there is no event (117) to be output, the block is an invalid block, so the code representing the invalid block is output to the transmission buffer (7) as encoded data (107).
また、第５図の例において例えば閾値を４または５に設
定したときの事象記憶部（１４）に記憶される事象と量
子化部（４）で量子化を行う係数の個数はそれぞれ次の
ようになる。Furthermore, in the example of FIG. 5, for example, when the threshold value is set to 4 or 5, the events stored in the event storage section (14) and the number of coefficients to be quantized by the quantization section (4) are as follows. become.
閾値４のときＱ（５）からＱ（８）で零係数が４つ連続
するため量子化打ち切りの条件を満たし、ゼロランと非
零係数値の組として記憶される事象は（０，２０）、
（２，１５）であり、量子化を行う変換係数の個数はＱ
（１）からＱ（８）までの８個となる。When the threshold is 4, there are four consecutive zero coefficients from Q(5) to Q(8), which satisfies the condition for quantization truncation, and the event stored as a set of zero run and non-zero coefficient value is (0, 20).
(2, 15), and the number of transform coefficients to be quantized is Q
There are eight items from (1) to Q(8).
閾値５のときＱ（１４）からＱ（１８）で零係数が５つ
連続するため記憶される事象は（０，２０）、 （２，
１５）、 （４、５）、 （３，２）であり、量子化を
行う係数の個数はＱ（１）からＱ（−１８）までの１８
個となる。When the threshold is 5, there are five consecutive zero coefficients from Q(14) to Q(18), so the events to be stored are (0, 20), (2,
15), (4, 5), (3, 2), and the number of coefficients to be quantized is 18 from Q(1) to Q(-18).
Become an individual.
先に述べたように一般に変換係数は低周波から高周波成
分になるに従って強度が弱くなるため、量子化した結果
の量子化係数Ｑ（ｉ） （１１３）もｊか大きくなるほ
ど連続して零となる確率か高い。As mentioned earlier, the intensity of the transform coefficient generally decreases from low to high frequency components, so the quantization coefficient Q(i) (113) as a result of quantization also becomes zero continuously as j increases. The probability is high.
従って、閾値（１１５）を小さくするほど係数の伝送範
囲が制限され量子化を要する係数の個数が少なくなると
同時に、発生する情報量も減少する。Therefore, the smaller the threshold value (115) is, the more the coefficient transmission range is restricted and the number of coefficients that require quantization is reduced, and at the same time, the amount of generated information is also reduced.
従って、閾値（１１５）の値を量子化ステップサイズと
同様にバッファ残量（↓０９）から適応的にフィードバ
ック制御すれば情報発生量の平滑化がより細かく可能に
なる。Therefore, if the value of the threshold value (115) is adaptively feedback-controlled from the remaining buffer amount (↓09) in the same way as the quantization step size, the amount of information generated can be smoothed more precisely.
また、本実施例によれば必ずしもすべての２次元変換係
数が必要でないため、１回目の１次元線形変換演算時も
２次元変換係数を求めるのに必要な１次元変換係数のみ
求めるようにしておけば変換係数を求めるための演算処
理量かさらに削減でき一層の効果が得られる。In addition, according to this embodiment, not all the two-dimensional transformation coefficients are necessarily required, so only the one-dimensional transformation coefficients necessary to obtain the two-dimensional transformation coefficients can be calculated during the first one-dimensional linear transformation calculation. In this case, the amount of arithmetic processing required to obtain the conversion coefficients can be further reduced, and further effects can be obtained.
［発明の効果」
以上のように、この発明によれば連続する零の量子化係
数の個数により、後続する変換係数を量子化し符号化す
るかどうかを判定するようにしたので、１次元線形変換
・量子化・有効無効識別・２次元可変長符号化を行うた
めの事象生成に要する演算量・処理時間を削減できる効
果がある。[Effects of the Invention] As described above, according to the present invention, it is determined whether or not to quantize and encode subsequent transform coefficients based on the number of consecutive zero quantization coefficients. - It has the effect of reducing the amount of calculations and processing time required for event generation for quantization, valid/invalid identification, and two-dimensional variable length encoding.
第１図は本発明の一実施例を説明するブロック図、第２
図は本発明の詳細な説明するフローチャート図、第３図
は従来例のブロック図、第４図は変換係数ブロックの性
質を説明するための図、第５図は符号の割当てを説明す
るための図である。
（１）はブロック化部、（２）は線形変換部、（３）は
スキャン変換部、（４）、（１４）は量子化部、（５）
は有効無効識別部、（６）は符号化部、（７）は送信バ
ッファ、（８）は符号化制御部、（９）、（１０）は１
次元線形変換部、（１，１）はゼロカウンタ、（１２）
は閾値設定部、（１３）は判定部、（１４）は事象記憶
部、（１５）は符号割当て部、（１１１）は１次元変換
係数ブロック、（１１２）は２次元変換係数、（１１４
）は計数値、（１１５）は閾値、（１，１６）は判定結
果、（１１−７）は事象である。
なお図中、同一符号は同一または相当部分を示す。
代 理 人 大 岩 増 雄＝
〜〜イ
第
２
図
第４図
水平方向周波酸分
２次元変換係数ブロックＦ（ｕ、ｖ）
＠小ξユ〈ごFIG. 1 is a block diagram explaining one embodiment of the present invention, and FIG.
3 is a block diagram of a conventional example, FIG. 4 is a diagram for explaining the properties of transform coefficient blocks, and FIG. 5 is a diagram for explaining code assignment. It is a diagram. (1) is a blocking section, (2) is a linear conversion section, (3) is a scan conversion section, (4), (14) is a quantization section, (5)
is a valid/invalid identifier, (6) is an encoder, (7) is a transmission buffer, (8) is an encoding control unit, (9) and (10) are 1
Dimensional linear transformation unit, (1, 1) is a zero counter, (12)
is a threshold setting section, (13) is a judgment section, (14) is an event storage section, (15) is a code assignment section, (111) is a one-dimensional transformation coefficient block, (112) is a two-dimensional transformation coefficient, (114) is a one-dimensional transformation coefficient block, (112) is a two-dimensional transformation coefficient,
) is a count value, (115) is a threshold value, (1, 16) is a determination result, and (11-7) is an event. In the drawings, the same reference numerals indicate the same or corresponding parts. Agent Masuo Oiwa=
～～A Fig. 2 Fig. 4 Horizontal direction frequency acid component two-dimensional conversion coefficient block F(u,v) @small ξyu〈go
Claims (1)
変換を行い変換領域で低域から高域へ変換係数を順次量
子化し符号化する変換符号化方式において、 ブロック化した入力信号系列に１次元線形変換を施し１
次元変換係数ブロックを得る手段と、前記１次元変換係
数ブロックにさらに直交する１次元線形変換を施し低域
から高域へ順次１つの２次元変換係数を得る手段と、 前記２次元変換係数を所定の量子化特性で量子化した量
子化係数のうち連続する零係数の個数を計数する手段と
、 前記量子化係数列から非零の係数と前記非零の係数が現
れるまでに前記計数手段により計数された連続零係数の
個数とを組としてブロック単位に記憶する手段と、 符号化情報発生量を所定の伝送情報量に近付けるために
符号化伝送する連続零係数の個数の閾値を送信バッファ
のデータ残量から設定する手段と、前記連続零係数を計
数した値が前記閾値を越えたとき後続する２次元変換係
数を求めるための１次元線形変換及び量子化処理を打ち
切り前記記憶された組毎に符号の割当てを行う手段とを
備えたことを特徴とする変換符号化方式。[Claims] In a transform encoding method in which an input signal sequence is divided into blocks, then two-dimensional linear transformation is performed, and transform coefficients are sequentially quantized and encoded from low to high frequencies in the transform domain. A one-dimensional linear transformation is applied to the input signal sequence.
means for obtaining a dimensional transform coefficient block; means for further performing orthogonal one-dimensional linear transform on the one-dimensional transform coefficient block to sequentially obtain one two-dimensional transform coefficient from a low range to a high range; means for counting the number of consecutive zero coefficients among the quantized coefficients quantized with the quantization characteristic; and counting by the counting means until a non-zero coefficient and the non-zero coefficient appear from the quantized coefficient sequence. a means for storing the number of continuous zero coefficients as a set in units of blocks; means for setting from the remaining amount, and when the counted value of the continuous zero coefficients exceeds the threshold value, the one-dimensional linear transformation and quantization processing for obtaining the subsequent two-dimensional transformation coefficients are terminated for each of the stored sets; 1. A transform encoding method, comprising: means for assigning codes.
Priority Applications (9)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2034658A JPH0822064B2 (en) | 1990-02-15 | 1990-02-15 | Transform coding method |
| US07/564,824 US5086488A (en) | 1989-08-19 | 1990-08-09 | Transform coding apparatus |
| DE69016880T DE69016880T2 (en) | 1989-08-19 | 1990-08-11 | Transformation coding device. |
| EP90115439A EP0414074B1 (en) | 1989-08-19 | 1990-08-11 | Transform coding apparatus |
| FI903989A FI98111C (en) | 1989-08-19 | 1990-08-13 | Transform coding apparatus |
| KR1019900012397A KR930009872B1 (en) | 1989-08-19 | 1990-08-13 | Changing coding apparatus |
| NO903624A NO303480B1 (en) | 1989-08-19 | 1990-08-16 | Transformation coding device |
| AU61069/90A AU622572B2 (en) | 1989-08-19 | 1990-08-16 | Transform coding apparatus |
| CA002023440A CA2023440C (en) | 1989-08-19 | 1990-08-16 | Transform coding apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2034658A JPH0822064B2 (en) | 1990-02-15 | 1990-02-15 | Transform coding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPH03238970A true JPH03238970A (en) | 1991-10-24 |
| JPH0822064B2 JPH0822064B2 (en) | 1996-03-04 |
Family
ID=12420542
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2034658A Expired - Lifetime JPH0822064B2 (en) | 1989-08-19 | 1990-02-15 | Transform coding method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH0822064B2 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7289673B2 (en) | 1998-11-30 | 2007-10-30 | Microsoft Corporation | Decoding macroblock type and coded block pattern information |
| US9077960B2 (en) | 2005-08-12 | 2015-07-07 | Microsoft Corporation | Non-zero coefficient block pattern coding |
| US9088785B2 (en) | 2001-12-17 | 2015-07-21 | Microsoft Technology Licensing, Llc | Skip macroblock coding |
-
1990
- 1990-02-15 JP JP2034658A patent/JPH0822064B2/en not_active Expired - Lifetime
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7289673B2 (en) | 1998-11-30 | 2007-10-30 | Microsoft Corporation | Decoding macroblock type and coded block pattern information |
| US9088785B2 (en) | 2001-12-17 | 2015-07-21 | Microsoft Technology Licensing, Llc | Skip macroblock coding |
| US9538189B2 (en) | 2001-12-17 | 2017-01-03 | Microsoft Technology Licensing, Llc | Skip macroblock coding |
| US9774852B2 (en) | 2001-12-17 | 2017-09-26 | Microsoft Technology Licensing, Llc | Skip macroblock coding |
| US10368065B2 (en) | 2001-12-17 | 2019-07-30 | Microsoft Technology Licensing, Llc | Skip macroblock coding |
| US9077960B2 (en) | 2005-08-12 | 2015-07-07 | Microsoft Corporation | Non-zero coefficient block pattern coding |
Also Published As
| Publication number | Publication date |
|---|---|
| JPH0822064B2 (en) | 1996-03-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR930009872B1 (en) | Changing coding apparatus | |
| EP0873018A1 (en) | Image encoding/decoding method, image encoder/decoder and image encoding/decoding program recording medium | |
| JPH02305182A (en) | Picture signal compressing and encoding device | |
| US20060251333A1 (en) | Image coding and decoding methods, image coding and decoding apparatuses, and recording media for image coding and decoding programs | |
| JPH07123269A (en) | Corder for picture signal | |
| Parisot et al. | EBWIC: A low complexity and efficient rate constrained wavelet image coder | |
| JP3603000B2 (en) | Huffman encoding apparatus, Huffman encoding method, and recording medium recording Huffman encoding processing program | |
| JPH06292019A (en) | Picture data compressor and picture code compressor | |
| JPH03238970A (en) | Conversion coding system | |
| US5724096A (en) | Video signal encoding method and apparatus employing inter-block redundancies | |
| US6084913A (en) | Method for compressing dynamic image information and system and device therefor | |
| JP2503706B2 (en) | Transform coding method | |
| JP2915922B2 (en) | Image signal compression coding device | |
| JPH0270127A (en) | Conversion encoding system | |
| JP2503698B2 (en) | Transform coding method | |
| JP2931328B2 (en) | Image signal compression coding device | |
| JP3230336B2 (en) | Recompression device and recompression system | |
| JP2633118B2 (en) | Image data compression apparatus and method | |
| JP3001758B2 (en) | Huffman encoding device and Huffman encoding method | |
| JP3191950B2 (en) | Image coding device | |
| JPH0575867A (en) | Image data coder | |
| JP3866539B2 (en) | Encoding method, decoding method, encoding device, decoding device, encoding program, decoding program, and program recording medium thereof | |
| JPH07222152A (en) | Picture processor and compressed data control method in the picture processor | |
| JPH04208775A (en) | Bit distribution encoding system | |
| JP2820807B2 (en) | Image data encoding method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20080304 Year of fee payment: 12 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20090304 Year of fee payment: 13 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20100304 Year of fee payment: 14 |
|
| EXPY | Cancellation because of completion of term |