JPH02248162A - Picture data encoding system - Google Patents
Picture data encoding systemInfo
- Publication number
- JPH02248162A JPH02248162A JP6767789A JP6767789A JPH02248162A JP H02248162 A JPH02248162 A JP H02248162A JP 6767789 A JP6767789 A JP 6767789A JP 6767789 A JP6767789 A JP 6767789A JP H02248162 A JPH02248162 A JP H02248162A
- Authority
- JP
- Japan
- Prior art keywords
- value
- difference
- gradation
- encoded
- predicted
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 claims description 16
- 238000006243 chemical reaction Methods 0.000 abstract description 8
- 230000000694 effects Effects 0.000 abstract description 8
- 238000013144 data compression Methods 0.000 abstract description 6
- 238000010924 continuous production Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 5
- 238000007906 compression Methods 0.000 description 4
- 230000006835 compression Effects 0.000 description 4
- 230000002411 adverse Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T9/00—Image coding
- G06T9/004—Predictors, e.g. intraframe, interframe coding
Abstract
Description
【発明の詳細な説明】
〔産業上の利用分野〕
本発明はファクシミリ、テレヒジョン等の画像データを
圧縮符号化するための画像データ符号化方式に関し、特
に予測符号化方式を用いて画像データの圧縮効果を図る
ようにするものである。[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an image data encoding method for compressing and encoding image data of facsimile, telephony, etc., and in particular to compression of image data using a predictive encoding method. It is intended to be effective.
画像入力装置から入力される画像データを圧縮して符号
化する高能率符号化方式として、予測符号化方式がある
。この方式はある画素の階調値をその周辺の符号化済み
画素の階調値に基づく予測関数を用いて予測し、予測値
と実際の階調値との差分値を予め定めた非線形特性にし
たがって量子化す不ものである。2. Description of the Related Art A predictive encoding method is a highly efficient encoding method for compressing and encoding image data input from an image input device. This method predicts the tone value of a certain pixel using a prediction function based on the tone values of encoded pixels around it, and then calculates the difference between the predicted value and the actual tone value using predetermined nonlinear characteristics. Therefore, it is impossible to quantize it.
この場合、非線形特性は主として差分値の出現頻度によ
っており、出現頻度の高い部分は差分値を細かい刻み幅
で場合骨けし、出現頻度の低い部分は逆に大まかな刻み
幅で場合分けし、それぞれの場合に対応する代表値を変
換差分値とするものである。差分値の出現頻度は一般に
その絶対値が小さなものほど高く、大きなものほど低く
なる傾向がある。In this case, the nonlinear characteristics are mainly determined by the frequency of appearance of the difference values, and the parts with high frequency of appearance are divided into small steps, and the parts with low frequency of appearance are divided into cases with coarse steps. The representative value corresponding to the case is set as the conversion difference value. Generally, the smaller the absolute value of the difference value, the higher the frequency of appearance of the difference value, and the larger the absolute value, the lower the frequency of appearance of the difference value.
第４図は差分値と変換差分値との変換特性を示す非線形
特性図である。この特性図によると、差分値の絶対値が
大きくなるほど１つの変換差分値に対応する差分値の変
域が大きくなることが分かる。例えば差分値の絶対値が
「２３」を超える場合は「５０」であろうとｒｌ ＯＯ
Ｊであろうと全て１２７」と変換されてしまう。このこ
とは、差分値の絶対値が大きくなるほど実際の差分値と
変換差分値との誤差が大きくなる可能性が高いことを示
している。この誤差は符号化した画像データを復号化す
る際に原画像の階調値と復元画像の階調値との誤差とな
って現れる。FIG. 4 is a nonlinear characteristic diagram showing the conversion characteristics between the difference value and the converted difference value. According to this characteristic diagram, it can be seen that as the absolute value of the difference value becomes larger, the range of the difference value corresponding to one converted difference value becomes larger. For example, if the absolute value of the difference value exceeds "23", even if it is "50", rl OO
All characters, even J, are converted to 127. This indicates that the larger the absolute value of the difference value, the higher the possibility that the error between the actual difference value and the converted difference value will become larger. This error appears as an error between the gradation value of the original image and the gradation value of the restored image when the encoded image data is decoded.
ところで、一般に差分値が大きくなるのは原画像の階調
値が激しく変化する部分、例えば画像の輪郭部分で起き
る。そこで前述のように差分値の絶対値が実際よりも小
さな値に変換されると、復元画像での階調数の変化が原
画像の階調数の変化よりもなだらかになり、輪郭部分に
ボケが生しることになる。Incidentally, the difference value generally becomes large in a portion where the gradation value of the original image changes drastically, for example, in the contour portion of the image. Therefore, as mentioned above, if the absolute value of the difference value is converted to a smaller value than the actual value, the change in the number of gradations in the restored image will be more gradual than the change in the number of gradations in the original image, and the outline will be blurred. will be born.
さらに予測符号化では既に符号化した画素の階調値をも
とに新たな画素を予測するので、いったん大きな誤差が
生じると、それが後の画素にまで悪影響を及ぼすことに
なる。特に画素における輪郭部というのは極めて重要な
情報を含んでいる部分であり、ここにボケが生じると画
質を著しく劣化させることになる。Furthermore, in predictive coding, new pixels are predicted based on the gradation values of pixels that have already been encoded, so once a large error occurs, it will adversely affect subsequent pixels. In particular, the contour portion of a pixel contains extremely important information, and if blurring occurs here, the image quality will be significantly degraded.
この点を改善するために、差分値が所定値を超える場合
と超えない場合とで場合分けし、それぞれ異なる処理を
施す方式が提案されている（例えば特開昭５６−１２９
４８２号＝ １階調データの圧縮方式Ｊ）。In order to improve this point, a method has been proposed in which cases are divided into cases in which the difference value exceeds a predetermined value and cases in which it does not exceed a predetermined value, and different processing is applied to each case (for example, Japanese Patent Laid-Open No. 56-129
No. 482 = 1-gradation data compression method J).
この方式は、被読取体から読取りされた階調を有する連
続する複数の階調データをコード化して圧縮する方式に
おいて、連続する階調データ間の階調の差分値が所定値
以下の場合はその差分に対応するコードでコード化し、
所定値以上の場合はその階調に対応するコードでコード
化するようにしている。これにより差分の少ない場合が
非常に多いことから、全体としてコード数を減少するこ
とができ、また階調の絶対値のコードをときどき出力す
るので、誤りコード発生後に誤りコードの波及を防止す
ることが出来る。This method is a method that encodes and compresses a plurality of consecutive gradation data having gradations read from an object to be read. Code it with the code corresponding to the difference,
If it is more than a predetermined value, it is encoded with a code corresponding to that gradation. Since there are many cases where the difference is small, the number of codes can be reduced overall, and since the code of the absolute value of the gradation is sometimes output, it is possible to prevent the error code from spreading after the error code occurs. I can do it.
ところが、この方式では、差分値が所定値以下の場合は
差分値をそのまま符号化しているため、原画像の階調数
が１６階階調度と少ない場合は圧縮効果が上がるが、２
５６階調程度になると絶対値が所定値以下の差分値を全
てそのまま符号化することはあまり圧縮効果が上がらな
い。前述の第４図において、絶対値が３５以下の差分値
を全てそのまま符号化するとそれだけで６ビツトのデー
タが必要となり、原画像の８ビツトデータと比べても圧
縮効果が上がっていない。However, in this method, if the difference value is less than a predetermined value, the difference value is encoded as is, so if the number of gradations in the original image is as small as 16 gradations, the compression effect increases, but 2
At about 56 gradations, encoding all the difference values whose absolute values are less than a predetermined value as they are does not improve the compression effect very much. In FIG. 4 mentioned above, if all the difference values whose absolute values are 35 or less are encoded as they are, 6-bit data is required, and the compression effect is not improved even compared to the 8-bit data of the original image.
本発明はデータ圧縮の効果を上げると共に、誤りコード
の連続発生を防止することを目的とする。It is an object of the present invention to improve the effectiveness of data compression and to prevent the continuous occurrence of error codes.
本発明は多階調の画像データを圧縮符号化する際に、あ
る画素の階調値をその周辺の符号化済み画素の階調値に
基づく予測関数を用いて予測し、その予測値と実際の階
調値との差分値をね分化する符号化方式において、上記
差分値の絶対値が所定値内のときは当該差分値を予め定
めた非線形特性に従って変換したのちその変換差分値を
符号化するようにし、上記差分値の絶対値が所定値を超
えるときは上記階調値をそのまま符号化するようにする
。また上記変換差分値を符号化する際に上記変換差分値
の等しい画素が連続する場合は、当該連続する画素を１
つのランとしてそのランレングスを符号化するようにす
る。When compressing and encoding multi-gradation image data, the present invention predicts the gradation value of a certain pixel using a prediction function based on the gradation values of encoded pixels surrounding it, and then compares the predicted value with the actual value. In an encoding method that divides the difference value between the gradation value of When the absolute value of the difference value exceeds a predetermined value, the tone value is encoded as is. In addition, when encoding the transform difference value, if there are consecutive pixels with the same transform difference value, the consecutive pixels are
The run length is encoded as one run.
符号化済み画素の階調値を用いて符号化画素の値を予測
し、その予測値と実際の階調値との差分値を求める。次
いで、この差分値の絶対値が所定値を超えるか否か判定
し、所定値を超えるときは階調値をそのまま符号化し、
所定値内のときは予め定めた非線形特性に従って変換差
分値を求め、この変換差分値を符号化する。The value of the encoded pixel is predicted using the gradation value of the encoded pixel, and the difference value between the predicted value and the actual gradation value is determined. Next, it is determined whether the absolute value of this difference value exceeds a predetermined value, and if it exceeds the predetermined value, the tone value is encoded as is,
When it is within a predetermined value, a transform difference value is obtained according to predetermined nonlinear characteristics, and this transform difference value is encoded.
出現頻度の高い所定値内の差分値は、変換差分値を符号
化するためデータ圧縮の効果が上がり、所定値を超える
出現頻度の低い差分値は、階調値をそのまま符号化する
ため変換誤差の波及を防止することが出来る。Difference values within a predetermined value that occur frequently are encoded as converted difference values, increasing the data compression effect, while differential values that occur less frequently than a predetermined value are encoded as tone values, resulting in conversion errors. It is possible to prevent the spread of
本発明による画像データ符号化方式は、符号化済み画素
Ａ、Ｂ、Ｃ，Ｄと符号化画素Ｘとが第１図に示す関係に
ある場合、まず符号化画素Ｘの値を符号化済み画素Ａ−
Ｄの階調値を用いて予測する。予測値父は次式で求めら
れる。In the image data encoding method according to the present invention, when encoded pixels A, B, C, D and encoded pixel X have the relationship shown in FIG. A-
Prediction is made using the gradation value of D. The predicted value is calculated using the following formula.
父−（５Ａ＋２Ｂ＋２Ｃ−Ｄ＞／８
ここでＡ−Ｄの値は実際の階調値そのままではなく、復
号をした際に得られる値である。つまり符号化時にロー
カル復号を行い、その値を用いて予測する。Father - (5A + 2B + 2C - D>/8 Here, the value of A - D is not the actual gradation value as it is, but the value obtained when decoding. In other words, local decoding is performed at the time of encoding, and that value is used. to predict.
次いで、予測値父と実際の階調値Ｘとの差分値ｅを求め
る。Next, the difference value e between the predicted value and the actual tone value X is determined.
ｅ＝Ｘ−父
続いて、この差分値ｅの値を第２図に示す非線形特性に
したがい変換する。差分値ｅの絶対値が３５以上の場合
は実際の階調値をそのまま符号化し、絶対値が３５未満
の場合は差分値を１３段階に分けてそれぞれ１つの変換
差分値に対応させ符号化する。e=X−Subsequently, the value of this difference value e is converted according to the nonlinear characteristics shown in FIG. If the absolute value of the difference value e is 35 or more, the actual gradation value is encoded as is, and if the absolute value is less than 35, the difference value is divided into 13 stages and encoded, each corresponding to one converted difference value. .
さらに差分値ｅの絶対値が３５未満でかつ変換差分値が
等しい画素が連続して現れる場合は、その連続した個数
をもとにランレングス符号化を施し、変換差分値を表す
符号列の後にランレングスを表す符号列が連なるという
形をとる。Furthermore, if pixels with the absolute value of the difference value e less than 35 and the same transform difference value appear consecutively, run-length encoding is performed based on the number of consecutive pixels, and after the code string representing the transform difference value, It takes the form of a series of code strings representing run lengths.
具体的な符号化例を第３図（ａｌ〜（Ｃ）に示す。図（
ａ）は変換差分値に対応する符号化例を示し、先頭ピン
トがフラグピント、残りビットが１−２符号である。一
般にＸ−２符号とは、ラン長が２×−１ビツト以下なら
ばＸビットで符号化を行い、ラン長がそれより長い場合
には、２ビツト（１ビツトはフラグビット）ずつ必要に
応じて符号を付加する方法で作られる符号化方式である
。Specific encoding examples are shown in Figure 3 (al to (C).
A) shows an example of encoding corresponding to the conversion difference value, in which the first focus is the flag focus and the remaining bits are the 1-2 code. In general, X-2 code encodes with X bits if the run length is 2 x -1 bits or less, and if the run length is longer than that, encodes with 2 bits (1 bit is a flag bit) as necessary. This is an encoding method that is created by adding a code.
図（ｂｌは階調値をそのまま符号化する例を示し、先頭
ビットがフラグビット、残りビットが２進数コードであ
る。例えば階調値「５ｏ」の符号化列はｒｌｏｏｌｌｏ
ｏｌｏＪとなる。Figure (bl shows an example in which the gradation value is encoded as it is, the first bit is a flag bit, and the remaining bits are binary codes. For example, the encoded string of gradation value "5o" is rloollo
It becomes oloJ.
図（Ｃ１はランレングスの符号化例を示す。例えば「３
」の符号化列はＩＱ １１Ｊ、「６」の符号化列はＩＱ
Ｏ１１０Ｊ、「１１」の符号化列は「０００１０１１
Ｊとなる。Figure (C1 shows an example of run length encoding. For example, “3
” is IQ 11J, and “6” is IQ
O110J, the encoded string of “11” is “0001011”
It becomes J.
本発明によれば、予測値と実際の階調値との差分値の絶
対値が所定値を超えたときは、階調値に対応した符号化
列に変換するので、大きな誤差が発生してそれが伝播し
ていくのを防ぐことができ、画像を復元した際に画質の
劣化を抑えることができ、ことに輪郭部分をはっきりと
再現させることができる。また、前記差分値の絶対値が
所定値を超えなかった場合は、当該差分値を予め定めた
非線形特性にしたがって変換し、その変換差分値に対応
した符号化列に変換するので、データ圧縮の効果を上げ
ることができる。According to the present invention, when the absolute value of the difference between the predicted value and the actual gradation value exceeds a predetermined value, it is converted into a coded string corresponding to the gradation value, so that a large error does not occur. It is possible to prevent this from propagating, suppressing deterioration in image quality when the image is restored, and in particular, making it possible to clearly reproduce outlines. In addition, if the absolute value of the difference value does not exceed a predetermined value, the difference value is converted according to predetermined nonlinear characteristics and converted into a coded string corresponding to the converted difference value, so data compression is improved. You can increase the effect.
また、変換差分値の等しい画素が連続する場合は該画素
列を１つのランとしてそのランレングスを符号化するの
で、データ圧縮の効果をさらに上げることができる。Further, when pixels having the same conversion difference value are consecutive, the pixel string is treated as one run and the run length is encoded, so that the effect of data compression can be further improved.
第１図は画素列を示す図、
第２図は本発明による非線形特性を示す図、第３図は本
発明による符号化例を示す表、第４図は従来の非線形特
性を示す図である。FIG. 1 is a diagram showing a pixel column, FIG. 2 is a diagram showing nonlinear characteristics according to the present invention, FIG. 3 is a table showing an encoding example according to the present invention, and FIG. 4 is a diagram showing conventional nonlinear characteristics. .
Claims (2)
画素の階調値をその周辺の符号化済み画素の階調値に基
づく予測関数を用いて予測し、その予測値と実際の階調
値との差分値を符号化する符号化方式において、 上記差分値の絶対値が所定値内のときは当該差分値を予
め定めた非線形特性に従って変換したのちその変換差分
値を符号化するようにし、上記差分値の絶対値が所定値
を超えるときは上記階調値をそのまま符号化するように
することを特徴とする画像データ符号化方式。(1) When compressing and encoding multi-gradation image data, the gradation value of a certain pixel is predicted using a prediction function based on the gradation values of encoded pixels surrounding it, and the predicted value and actual In an encoding method that encodes the difference value between the gradation value of An image data encoding method characterized in that when the absolute value of the difference value exceeds a predetermined value, the gradation value is encoded as is.
の等しい画素が連続する場合、当該連続する画素を１つ
のランとしてそのランレングスを符号化することを特徴
とする請求項１記載の画像データ符号化方式。(2) When encoding the transform difference value, if pixels having the same transform difference value are consecutive, the run length is encoded with the consecutive pixels as one run. image data encoding method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP6767789A JPH02248162A (en) | 1989-03-22 | 1989-03-22 | Picture data encoding system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP6767789A JPH02248162A (en) | 1989-03-22 | 1989-03-22 | Picture data encoding system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH02248162A true JPH02248162A (en) | 1990-10-03 |
Family
ID=13351867
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP6767789A Pending JPH02248162A (en) | 1989-03-22 | 1989-03-22 | Picture data encoding system |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH02248162A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7715638B2 (en) | 2003-01-13 | 2010-05-11 | Nokia Corporation | Processing of images using a limited number of bits |
| JP2011040834A (en) * | 2009-08-06 | 2011-02-24 | Ricoh Co Ltd | Image processing apparatus and image processing method |
| JP2011044869A (en) * | 2009-08-20 | 2011-03-03 | Ricoh Co Ltd | Image processor, and method of image processing |
-
1989
- 1989-03-22 JP JP6767789A patent/JPH02248162A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7715638B2 (en) | 2003-01-13 | 2010-05-11 | Nokia Corporation | Processing of images using a limited number of bits |
| USRE43256E1 (en) | 2003-01-13 | 2012-03-20 | Nokia Corporation | Processing of images using a limited number of bits |
| JP2011040834A (en) * | 2009-08-06 | 2011-02-24 | Ricoh Co Ltd | Image processing apparatus and image processing method |
| US8515209B2 (en) | 2009-08-06 | 2013-08-20 | Ricoh Company, Limited | Image processing apparatus and method for image processing |
| JP2011044869A (en) * | 2009-08-20 | 2011-03-03 | Ricoh Co Ltd | Image processor, and method of image processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4343440B2 (en) | Real-time algorithm and architecture for encoding images compressed by DWT-based techniques | |
| JPH0485621A (en) | Rounding device | |
| US5719961A (en) | Adaptive technique for encoder and decoder signal transformation | |
| KR960700609A (en) | Moving picture encoding method, moving picture decoding method, moving picture recording medium and moving picture encoding apparatus (animation encoding method, animation decoding method, animation recording medium and animation encoder) | |
| KR100496774B1 (en) | Method and apparatus for recompressing with data efficient quantization table for a digital video signal processor | |
| US5966470A (en) | Coding apparatus for image compression | |
| JPH02248162A (en) | Picture data encoding system | |
| JP2002064821A (en) | Method for compressing dynamic image information and its system | |
| JPH03140074A (en) | Moving picture coding device | |
| JPH08275153A (en) | Image compressor and image decoder | |
| JPH02131671A (en) | Picture data compression method | |
| JPH01129589A (en) | Image encoding system | |
| JPH02305271A (en) | Picture data compressing method | |
| JP2002077920A (en) | Picture compressing apparatus and method therefor | |
| JP2000013797A (en) | Image compressing device and image expanding device | |
| KR100303744B1 (en) | Method and device for compressing and expanding image | |
| JPH0621828A (en) | Vector quantizing decoder | |
| JP3102323B2 (en) | Image coding device | |
| JPH06125542A (en) | High efficiency encoding system | |
| JPH0629861A (en) | Data compression method | |
| JP3480644B2 (en) | Image encoding device and image decoding device | |
| JP2003299120A (en) | Color image compressing method | |
| JPH0352386A (en) | Picture data compression method | |
| KR100234239B1 (en) | Method and apparatus of quantizing for decreasing blocking effect | |
| JPH0816192A (en) | Data compression device |