JP2595291B2 - Three-dimensional image conversion coding device - Google Patents
Three-dimensional image conversion coding deviceInfo
- Publication number
- JP2595291B2 JP2595291B2 JP6134688A JP6134688A JP2595291B2 JP 2595291 B2 JP2595291 B2 JP 2595291B2 JP 6134688 A JP6134688 A JP 6134688A JP 6134688 A JP6134688 A JP 6134688A JP 2595291 B2 JP2595291 B2 JP 2595291B2
- Authority
- JP
- Japan
- Prior art keywords
- image data
- memory
- dimensional
- discrete sine
- predictive
- 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.)
- Expired - Fee Related
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Description
本発明は画像を変換符号化し、画像データを圧縮する
ための３次元画像変換符号化装置に関するものである。The present invention relates to a three-dimensional image transform coding apparatus for transform coding an image and compressing image data.
画像の画素間相関を利用して高能率な符号化を行う方
法としては本発明者等が既に特願昭62−240455号と提案
している外挿予測離散サイン変換による方法がある。As a method of performing highly efficient encoding using the correlation between pixels of an image, there is a method of extrapolation predictive discrete sine transform already proposed by the present inventors in Japanese Patent Application No. 62-240455.
しかしながらこの方法は画像データ１枚中の画素間相
関を利用しているので連続した画像データの変換に際し
ては画像データの時間的相関を利用することがなく非能
率的である。 本発明は上記の問題点に鑑みて発明したものであっ
て、その目的とするところは連続した画像データに対し
て時間的相関を利用してさらに高能率な符号化を行うこ
とができる３次元画像変換符号化装置を提供するにあ
る。However, since this method uses the correlation between pixels in one image data, it is inefficient when converting continuous image data without using the temporal correlation of the image data. SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and has as its object to perform three-dimensional encoding that can perform more efficient encoding on continuous image data by using temporal correlation. An image transformation coding apparatus is provided.
本発明は連続した複数枚の画像データを蓄積できるメ
モリと、このメモリに記憶した画像データのアクセス順
序を１画面の水平、垂直方向の所定画素数ずつか、水平
または垂直方向と時間方向の所定画素数ずつかに切り換
えることができるメモリ制御手段と、２次元外挿予測−
離散サイン変換により画像データの符号化を行う２次元
外挿予測離散サイン変換手段とを備えている。The present invention provides a memory capable of storing a plurality of continuous image data, and an access order of the image data stored in the memory is determined by a predetermined number of pixels in a horizontal and vertical direction of one screen, or by a predetermined number in a horizontal or vertical direction and a time direction. Memory control means that can switch between pixel counts and two-dimensional extrapolation prediction
A two-dimensional extrapolation predictive discrete sine transform means for encoding image data by discrete sine transform.
而して本発明装置ではメモリ制御手段により各枚の画
像データから水平、垂直方向の所定画素のブロックを切
換抽出して２次元外挿予測離散サイン変換手段により予
測変換符号化を行い、この予測変換符号化したデータを
メモリに蓄積した後にデータの内所定枚の画像データに
亘る所定画素数のブロックをメモリ制御手段により切換
抽出して２次元外挿予測離散サイン変換手段にて時間軸
方向の圧縮を行うのである。In the apparatus of the present invention, a block of predetermined pixels in the horizontal and vertical directions is switched and extracted from each image data by the memory control means, and predictive transform coding is performed by the two-dimensional extrapolation predictive discrete sine transform means. After the transform-encoded data is stored in the memory, a block of a predetermined number of pixels over a predetermined number of image data of the data is switched and extracted by the memory control means, and the two-dimensional extrapolated discrete sine transform means extracts the block in the time axis direction. It does compression.
以下本発明を実施例の回路構成により説明する。 第１図は一実施例の構成を示しており、画像データを
Ｎ＋１枚分蓄積できるメモリ１と、後述の２次元外挿予
測離散サイン変換手段３と、上記メモリ１に蓄積された
画像データから２次元外挿予測離散サイン変換手段３に
データを渡す際に、画像データ１画面内の水平、垂直方
向のＮ×Ｎ画素のブロックを渡すか、画像データの水平
または垂直方向と時間方向のＮ画素ずつを渡すかを切り
換えるメモリ制御手段２とから構成される。 ２次元外挿予測離散サイン変換手段３は第２図に示す
ように画像をブロックに分散してブロックの外の境界値
からブロック内の画像信号を外挿予測し、この予測によ
って得られた外挿予測信号をブロック内の原画像信号よ
り減じて外挿予測誤差信号として出力する２次元外挿予
測手段3aと、この外挿予測誤差信号を２次元離散サイン
変換する変換手段3bと、変換手段3bで変換された変化係
数をある閾値で切り捨て、更にある量子化ステップ巾に
よって量子化する量子化手段3cと、この量子化手段3cの
出力を逆量子化し２次元逆離散サイン変換して上記外挿
予測信号との加算を行うことで、次のブロックとの境界
値を再生する逆変換手段3dとをからなるものであり、画
像を予測した後に変換符号化するので、特に従来のDCT
方式に比べてブロックサイズを小さくすることができと
いう特徴がある。 而して第１図において連続した画像データＮ＋１枚を
メモリ１に蓄積する。次にメモリ制御手段２によってメ
モリ１に蓄積した画像データを第３図のように１枚ずつ
Ｎ×Ｎ画像のブロックに分け、２次元外挿予測離散サイ
ン変換手段３に入力し、予測変換符号化する。予測変換
符号化したデータは元の画素データのところのメモリ蓄
積する。 このようにしてＮ＋１枚の画像データをそれぞれ独立
に予測変換符号化し終えた後、第３図のように画像デー
タの予測変換符号化したデータのＮ枚に亘るＮ×Ｎ画素
のブロックをメモリ制御手段２により選択して２次元外
挿予測離散サイン変換手段３に入力し、ここで時間軸方
向の圧縮を行う。この時第４図のように画像データの垂
直方向と時間方向を選択することもでき、また第５図の
ように画像の水平方向と時間方向に選択することもでき
る。ここで一番上の画像データ部分は時間軸方向の境界
値として使用する。尚Ｎとしては４が実際的な値であ
る。 第６図は動画像の予測変換符号化に用いた実施例の回
路構成を示しており、第５図のようにＮ枚分のメモリ1
a,1bを２つ準備し、このメモリ1a,1bの入出力を切換手
段4a,4bで切換えるようになっている。 而して動画像の最初の１枚はメモリ1aのＮ枚目の画像
データ領域に入力し、続きの画像データをＮ枚分メモリ
1bに入力し、メモリ1aに蓄積した上記Ｎ枚目の画像デー
タを境界値として２次元外挿予測離散サイン変換手段３
にて予測変換符号化を行い、メモリ1b側の予測変換符号
化が実行されている間に最初からＮ−１枚目までの画像
データをメモリ1aに入力する。メモリ1b側の画像データ
の予測変換符号化が終了した後にメモリ1aのＮ枚目の画
像データをメモリ1bに入力し、このＮ枚目の画像データ
を境界値としてメモリ1a側の予測変換符号化を開始す
る。以下この手順を繰り返すことによって動画像の予測
変換符号化が可能となる。この時Ｎ枚の予測変換符号化
は動画像Ｎ−１枚分の時間内に終了することが必要とな
る。Hereinafter, the present invention will be described with reference to a circuit configuration of an embodiment. FIG. 1 shows a configuration of one embodiment, in which a memory 1 capable of storing N + 1 image data, a two-dimensional extrapolation predictive discrete sine transform unit 3 described later, and When passing the data to the two-dimensional extrapolation discrete sine transform means 3, a block of N × N pixels in the horizontal and vertical directions within one screen of the image data is passed or N or N pixels in the horizontal or vertical direction and the time direction of the image data are passed. And memory control means 2 for switching whether to deliver each pixel. The two-dimensional extrapolation discrete sine transform means 3 extrapolates the image signal in the block from the boundary values outside the block by dispersing the image into blocks as shown in FIG. A two-dimensional extrapolation predicting means 3a for subtracting the extrapolated prediction signal from the original image signal in the block and outputting it as an extrapolated prediction error signal; a converting means 3b for performing two-dimensional discrete sine transform on the extrapolated prediction error signal; The change coefficient converted in 3b is truncated at a certain threshold, and further quantized by a certain quantization step width. An output of the quantization means 3c is inversely quantized and subjected to two-dimensional inverse discrete sine transform to obtain And an inverse transform means 3d for reproducing a boundary value with the next block by performing addition with the interpolated prediction signal.
There is a feature that the block size can be reduced as compared with the system. In FIG. 1, N + 1 consecutive image data are stored in the memory 1. Next, the image data accumulated in the memory 1 by the memory control means 2 is divided into N × N image blocks one by one as shown in FIG. Become The data subjected to the predictive transform coding is stored in the memory at the original pixel data. After the N + 1 image data are independently predictively transformed and coded, the N × N pixel blocks of N data of the predictively coded data of the image data are subjected to memory control as shown in FIG. The data is selected by the means 2 and input to the two-dimensional extrapolated predictive discrete sine transform means 3, where compression in the time axis direction is performed. At this time, the vertical direction and the time direction of the image data can be selected as shown in FIG. 4, and the horizontal direction and the time direction of the image can be selected as shown in FIG. Here, the top image data portion is used as a boundary value in the time axis direction. Note that 4 is a practical value for N. FIG. 6 shows a circuit configuration of an embodiment used for predictive transform coding of a moving image. As shown in FIG.
Two a and 1b are prepared, and the input and output of the memories 1a and 1b are switched by the switching means 4a and 4b. Thus, the first image of the moving image is input to the Nth image data area of the memory 1a, and the subsequent image data is stored in the Nth image data area.
1b, and the N-th image data stored in the memory 1a is used as a boundary value for the two-dimensional extrapolated discrete sine transform means 3
Performs predictive transform coding, and inputs the N-1th image data from the beginning to the memory 1a while the predictive transform coding on the memory 1b side is being executed. After the predictive transform coding of the image data in the memory 1b is completed, the Nth image data in the memory 1a is input to the memory 1b, and the Nth image data is used as a boundary value in the predictive transform coding in the memory 1a. To start. Hereinafter, by repeating this procedure, predictive transform coding of a moving image becomes possible. At this time, it is necessary that the N predictive transform coding be completed within a time corresponding to N-1 moving images.
本発明は上述のように構成してあるから、画像データ
の時間経過に関する相関を利用て時間軸方向にも圧縮を
行うことができ、２次元外挿予測離散サイン変換よりも
圧縮率を上げることができる。Since the present invention is configured as described above, it is possible to perform compression also in the time axis direction using the correlation with respect to the lapse of time of image data, and it is possible to increase the compression ratio compared with the two-dimensional extrapolation predictive discrete sine transform. Can be.
第１図は本発明の一実施例の構成図、第２図は２付言外
挿予測離散サイン変換手段の構成図、第３図は画像デー
タの２次元外挿予測のためのメモリ走査手順を示す図、
第４図は外挿予測を時間軸方向に行う際のメモリ走査の
手順を示す図、第５図は第４図の別の走査方法を示す
図、第６図は本発明の別の実施例の構成図である。 １はメモリ、２はメモリ制御手段、３は２次元外挿予測
離散サイン変換手段である。FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a block diagram of a discrete sine transform means for two-part extrapolation prediction, and FIG. 3 is a memory scanning procedure for two-dimensional extrapolation prediction of image data. Diagram,
FIG. 4 is a diagram showing a procedure of memory scanning when extrapolation prediction is performed in the time axis direction, FIG. 5 is a diagram showing another scanning method of FIG. 4, and FIG. 6 is another embodiment of the present invention. FIG. 1 is a memory, 2 is a memory control means, and 3 is a two-dimensional extrapolation predictive discrete sine transform means.
───────────────────────────────────────────────────── フロントページの続き (72)発明者 山根 延元 岡山県岡山市津島中１丁目３番ＲＢ棟 103 (56)参考文献 特開 昭60−12884（ＪＰ，Ａ) 電子通信学会論文誌Ｊ69−Ｂ［７ ］（1986）Ｐ．686−697 ────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Nobumoto Yamane 1-3-3 RB Building, Tsushimanaka, Okayama City, Okayama Pref. -B [7] (1986) P. 686-697
Claims (1)
メモリと、このメモリに記憶した画像データのアクセス
順序を１画面の水平、垂直方向の所定画素数ずつか、水
平または垂直方向と時間方向の所定画素数ずつかに切り
換えることができるメモリ制御手段と、２次元外挿予測
−離散サイン変換により画像データの符号化を行う２次
元外挿予測離散サイン変換手段とを備え、メモリ制御手
段により各枚の画像データから水平、垂直方向の所定画
素数のブロックを切換抽出して２次元外挿予測離散サイ
ン変換手段により予測変換符号化を行い、この予測変換
符号化したデータをメモリに蓄積した後にデータの内所
定枚の画像データに亘る所定画素数のブロックをメモリ
制御手段により切換抽出して２次元外挿予測離散サイン
変換手段にて時間軸方向の圧縮を行うことを特徴とする
３次元画像変換符号化装置。A memory capable of storing a plurality of continuous image data, and an access order of the image data stored in the memory is determined by a predetermined number of pixels in a horizontal and vertical direction of one screen, or in a horizontal or vertical direction and a time direction. And a two-dimensional extrapolation discrete sine transform means for encoding image data by two-dimensional extrapolation / discrete sine transform. The memory control means A block having a predetermined number of pixels in the horizontal and vertical directions is selectively extracted from each image data, and subjected to predictive transform encoding by means of a two-dimensional extrapolated predictive discrete sine transform means, and the predictive transform-encoded data is stored in a memory. Thereafter, a block having a predetermined number of pixels over a predetermined number of image data in the data is switched and extracted by the memory control means, and the block is extracted by the two-dimensional extrapolation discrete sine transform means. Performing the direction of the compressed three-dimensional image transform coding apparatus according to claim.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP6134688A JP2595291B2 (en) | 1988-03-14 | 1988-03-14 | Three-dimensional image conversion coding device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP6134688A JP2595291B2 (en) | 1988-03-14 | 1988-03-14 | Three-dimensional image conversion coding device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPH01233892A JPH01233892A (en) | 1989-09-19 |
| JP2595291B2 true JP2595291B2 (en) | 1997-04-02 |
Family
ID=13168481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP6134688A Expired - Fee Related JP2595291B2 (en) | 1988-03-14 | 1988-03-14 | Three-dimensional image conversion coding device |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP2595291B2 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5327242A (en) * | 1993-03-18 | 1994-07-05 | Matsushita Electric Corporation Of America | Video noise reduction apparatus and method using three dimensional discrete cosine transforms and noise measurement |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU570439B2 (en) * | 1983-03-28 | 1988-03-17 | Compression Labs, Inc. | A combined intraframe and interframe transform coding system |
-
1988
- 1988-03-14 JP JP6134688A patent/JP2595291B2/en not_active Expired - Fee Related
Non-Patent Citations (1)
| Title |
|---|
| 電子通信学会論文誌Ｊ６９−Ｂ［７］（１９８６）Ｐ．６８６−６９７ |
Also Published As
| Publication number | Publication date |
|---|---|
| JPH01233892A (en) | 1989-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5386234A (en) | Interframe motion predicting method and picture signal coding/decoding apparatus | |
| CA2108778C (en) | Process for coding picture signals | |
| KR100209793B1 (en) | Apparatus for encoding/decoding a video signals by using feature point based motion estimation | |
| JP2900998B2 (en) | Block interpolation predictive encoding apparatus, decoding apparatus, encoding method and decoding method | |
| JPH0137065B2 (en) | ||
| EP0534282A1 (en) | A method for compressing a digitally coded video image | |
| JP2595291B2 (en) | Three-dimensional image conversion coding device | |
| JP2947389B2 (en) | Image processing memory integrated circuit | |
| JPH0514735A (en) | Picture processor | |
| JP3804745B2 (en) | Frame structure / field structure switching type image coding apparatus | |
| JPS62188565A (en) | Method and apparatus for picture data quantization | |
| JP2004521547A (en) | Video encoder and recording device | |
| US5475432A (en) | Hibrid video signal encoder having a block rearrangement capability for better compressibility | |
| JP2722353B2 (en) | Image signal encoding device | |
| JP2962537B2 (en) | Scanning method and apparatus using energy distribution of two-dimensional data | |
| JP2537246B2 (en) | Image coding method | |
| JP2795100B2 (en) | Image compression circuit and image decompression circuit | |
| JPH07123412A (en) | Moving image compression-encoding method | |
| JP2839055B2 (en) | Image editing device | |
| JP2687417B2 (en) | Automatic linear density switching system for facsimile machines | |
| JPH05292330A (en) | Picture data coding method | |
| JPH0687578B2 (en) | Image coding device | |
| JPH0851628A (en) | Moving picture processor, its control method and moving picture transmission system | |
| JPH04269084A (en) | Image converter | |
| JPH06233271A (en) | Moving picture decoding method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| LAPS | Cancellation because of no payment of annual fees |