1321953 VIT05-0273 18607twf.doc/g 九、發明說明： 【發明所屬之技術領域】 本發明是有關於一種縮放電路及其方法，且特別是有 關於一種採遞迴方式（recursive)之晝面縮放電路及其方 法。 【先前技術】1321953 VIT05-0273 18607twf.doc/g IX. Description of the Invention: [Technical Field] The present invention relates to a scaling circuit and a method thereof, and more particularly to a face rescaling method (recursive) Circuits and methods. [Prior Art]
在影像播放方面’為了適用不同之觀賞使用者及各相 異性質之影像，在同一種顯示器例如電視，往往採用不同 顯示影像規格。一般來說，常用之顯示影像規格為Nati〇n Television System Committee ( NTSC )及孙脱 Aitemati〇n by Line (PAL)。但在此兩系統中，掃描線數之水平掃描 頻率及垂直掃描頻率各不相同。同時在NTSC及pAL系統 中’有不同之亮度（luminance)及彩度(chr〇mi麵㈣規格， 因此在不同影像規格之轉換需透過晝面處理電路。 口旦面、縮放電路100之電路圖。原影像所使 用之資料傳輸線U〜L5,而透過晝 ，換之方式為將資料傳輪線u直接輸出，並將資 f L2上之·丨_存， 單 = 之電壓，以蚊比例組合輸 ^透過^fH4上之說經過緩衡11 iG2儲存， 再透過6十异早凡112與資料傳 =出至資料傳，，於是此畫面 r。。1 1〇2及计异早元ln、112以完成畫面縮小之功 5 1-321953 VIT05-0273 18607twf.doc/g 能。由上述得知，這樣會造成書面始 ^ , 、战旦囟細放電路100之設計缺 之彈性。 若要產生縮小_ 5]之觀晝 面縮放電路100。一般來說雷抓舛4加 而里新汉砟旦 〇口- 、门士地并工兄而5又6十4個緩衝器及1個計算 早凡’以㈣加〜並平均五條㈣傳輪 如爾輸線一之資料傳輸線的=例 V13、V14及V15 ,可得輸出電壓值為 v\\ + F12 + V13 + VIA + 厂15 5 如此-來，此設計之電路不但僅能 且當元件數量過多’將導致生產成本提高。1於^;數: 耦接關係複雜，會降低生產流程之可靠度。由於影H 具多樣性’並依顯不器及根據影像内容做調整，、羽 之晝面處理電路具有元件過多且缺乏彈性等缺點 【發明内容】 ' … 本發明的目的就是在提供-種晝面縮放電路，由 取遞迴（recursive)方式處理影像資料之結構，— 少電路面積，並可降低生產成本及提升良率，並二, 種不同影像轉換及加入新的影像規格。 、’。配合多 本發明的目的就是在提供一種晝面縮放方去 電路面積及生產成本，並可提升良率，更可配合 影像轉換及加入新的影像規格。 夕不同 計 本發明提出一種晝面縮放電路，包括一緩衝哭、 I所要轉 算單元及-控制單it。㈣單元減原影像規格二 6 VIT05-0273 18607twf.doc/g 換成的影像規格間的關係，以決定控制訊號的内容並發出 ^制訊紅緩衝ϋ。緩_接收此控制訊號而決定是否 ^出影像列儲存為緩衝影像列，及決定是否輸出所儲存 二衝衫像列。计异單元之—端祕至此緩衝器 器：輸出之緩衝影像列，並接收輸入影像列以對 列與緩衝影像列進行預設計算，而產生前述的輸 括儲:本發日^的較？實施例所述之晝面縮放電路’更包 所需的轉i方式健用以健存在轉換不同影像規格時 時，先至储财元決定控制訊號的内容 式，控制單元再㈣;的不同影像規格間的轉換方 將此控制訊號傳^至方式來決f空制訊號的内容而 憶體。 、’裔。此儲存單元包括非揮發性記 本發明另提出—錄 訊號而決定是否將^旦^放方法，包括：根據-控制 是否輸出所儲存為緩衝影像列及決定 此外，並接收—輪入旦1，接收從該緩衝影像列， 像列進行-預設計曾=，以對該輸入影像列與該緩衝影 訊號予緩衝Γ ^查詢儲存單元而發出控制 及決定由料單科料元之輪入端， 出而輸入至緩衝器的遽迴結構。可 1321953 VIT05-0273 18607twf.doc/g 減少電路面積、降低生產成本、提、， 不同存單元加入‘規:可配合多種 為讓本發明之上述和其他 易懂’下文特舉較佳實施例，並配合顯 明如下。 π附圖式，作詳細說 【實施方式】 圖2為本發明實施例之一畫面 塊圖。晝面縮放電路2。。包括—緩衝器2。、二電路方 一控制單元22及一儲存單元23。儲存星1早兀21、 像規格時所需轉換之方式，此影像規格二，存轉換影 像規格產生。儲存單元23並不==因=^不同影 可儲存多種轉換方式以提供至控制單方式， 22輸出控制訊號，當控制單元22要^用。控制單元 號之内容時，可先謂存單 ^ ㈣之控制訊 之韓換方4查珣所需要之影像規格間 之轉換方式，再_此轉财絲蚊 f制皁元22輸出控制訊號至緩衝器2〇衝：ς ; 是否將㈣ 以决疋疋否輸出所儲存之緩衝影 單元21之一端耦接至緩 。十'早兀21。计异 出之緩衝影像列，並接收輸旦 故衝$ 20所輸 缓衝影像列進行預設，以對輸入影像列與 中，儲存輸出影像列。在此實施例 ^ 可為一非揮發性記憶體。 衝哭tn由控龍餘制，將輪出影制儲存至緩 口口 ’再輸出此緩衝影像列至計算單元21與輸入影像 8 1321953 V1T05-0273 18607twf.doc/g 列再做運算之遞迴方式。可有 執行-適當加權處理。例如傳輪線之值 時，圖3為儲存單元縮小比為 ... —曲細小比為5 : 1白ij斜日刀主 言月同日守參考圖3及圖2，圖3之對照表第i 、十、表二 衝！§' 20儲存輸出影像列。第2〜4 ，、、扣不、是 遞迴，在每-次遞迴中，緩衝器20^前儲::::三:欠 列輸入至計算單元21以與新的輪i像 =裘_ 20再儲存經計算後之輸出 =動 時’表示將先輸出影像列輪人至ft 元21以執行第四次計算動作， 至计#早 後之輸出影像列。此時計算單元二 出影像值。輸出訊號電壓之關係可例為斤而之輸 輸電壓V2卜V22、V23、V24及5日守序之資料傳 K21 + V22 -厂 23 1-+ V24 2 +V25 ，輪出電壓可為 1/01 丄]/00 2 以 此遞迴方式，並配合儲存單元23之不_ 各影像規格之間的轉換。同時，儲存單元5對照表，可達到 規格經試驗後之更新最佳對照表，及輕 可儲存各影像 再者，畫面縮放單元200只使用緩衝、新的衫像規格。 久何态2〇及呌管OD __ 執行遞迴方式之計算，可節省電路元件及面uT，早凡21 單元200並可使用於垂直及水平方向 積。晝面縮放 <轉換。 圖4為工業影像規格YC万Cr 7422之部分像素點分佈 9 叫 1953 VIT05-0273 18607twf.d〇c/g 圖。圖5為工業影像規格YC/5Ct 420之部分像素點分佈 圖。請同時參考圖4及圖5，像素點401、403及501含有 亮度資料及彩度資料（亮度資料以空心圓圈符號表示，彩 度資料以點符號表示），像素點4〇2、404及502含有彩度 資料。像素點501之亮度資料傳輸電壓可由圖2之畫面縮 放單元200計算像素點4(Π、402、403、404之亮度資料傳 輸電壓得知’例如像素點4〇1、402、403、404之亮度資料 傳輸電壓為 V401、V402、V403、V404 (V403、V404 量 K40U F402 ~2 笮氕f ί)，則像素點501之亮度資料傳輸電壓為 + V403 2 + V404 ::素因無亮度資料，故亮度資料傳輸電壓量值視 5 I轉：方2 5之其他像素點可依此類推’但圖4至圖 5之轉換方式不限於四像素回 意數目之像素點執行遞迴計^像^之轉換，可為任 圖5之轉換經由晝面縮放面轉換。由圖4至 影像資料差異。 以遞迴方式計算以彌補 圖6為本發明另一實施例之全 圖。影像所錢之水平資料胁^面㈣魏之電路方塊 縮放電路轉換影像，僅使二^n〜L!5 ’而透過晝面 控繼：及放一一計算單元 換影像規格時所需轉換之方 儲存單元63儲存轉 *影像規格為ϋ應於兩不 10 VIT05-0273 18607twf.doc/g 同影像規袼產生。控制單A 62輸出控制訊號，當控 22要決定所輸出之控制訊號之内容時，可先至儲存單元63 f詢所f要之影像規格間之轉換方式，再根據此轉換方式 =定控制訊號之内容。控制單以2輸出控制訊號至緩衝 為60 ’而緩衝器6〇可根據此控制訊號選擇是否將輪 衝f像列，以蚊是否輸出所儲存之緩衝^ 象。十异早7L01。計算單元61之一端耦接至緩衝界2〇, 所輸出之緩衝影像列，並接收輸入影像 列，以對資料傳輸線L11〜L15之輸入影像列與緩衝影像 列進饤預設計算秘資料傳輸線L16輸出影像列。在此 施例中’儲存單元63可為一非揮發性記憶體。於是此畫面 ，放電路以儲存單元63、控制單元62、緩衝器60及；；算 r Μ來完成晝面縮小之功能。本實施例以水平傳輸線二 歹’以目同原理’垂直傳輸線亦可使用相同遞迴方式來 到畫面縮放之功能。本實施例以畫面縮小為例，但孰知此 技蟄=士當知，晝面放大亦可使用相同結構，所差別在 於計异單元所輸人/輸出各自之⑽傳輸線的多寡。 圖7為本發明實施例之晝面縮放方法之流^圖。該竺 面縮放方法，包括以下步驟：首先，在步驟s7〇i中，根 據-，制訊號而決定是否將一輸出影像列儲存為緩衝影像 ^。出影像列儲存為緩衝影像列，決定是否輪出所儲 存之緩衝影像列。在步驟S7G3巾，接收緩衝影像列 接收-輸人影像列，以對輸人影像列與緩衝影像列進行二 預设计异，並產生-輸出影像列。在步驟S7〇5中，健存 1-321953 VIT05-0273 18607twf.doc/g 在轉換影像規格時所需的一轉換方式，， 於兩不同影像規格間的關係所產生，當要〜式係因應 的内容時，先查詢所需的影像規格間的轉換^控制訊號 此轉換方絲蚊該控制訊號㈣容 、，^ ’再根據 中’根據原影像規格與所要轉換成的 步驟S707 決定該控制訊號的内容。此4面縮放 1間的關係， 非揮發性記憶體來儲存在轉換影像規格所^含^一 式。 才^而的—轉換方 絲上所述，在本發明之晝面縮放電路及复 使用控制單元提供控制訊號給緩衝器，由緩衝〜，由於 迴結構，使得電路元件及面積大幅減:且IS ，產成本及提升生產流程之可靠度。且配合儲 二球 :多種對照表’以增加處理影像規袼之彈性及達到：二二 換效能。由於採取遞迴及達到取佳轉 減少電路面積’降低生產成本及提升良率， 、'了酉己δ夕種不同影像轉換及加入新的影像規格。 和，火何熟習此技勢者，在不脫離本發明之精神 ε *可作些許之更動與潤飾，因 f圍當顯咐之申請專利範園所界定者為準J之保護 【圖式簡單說明】 3為¥知晝面縮放電路1 〇〇之電路圖。 圖2為本發明實施例之畫面縮放電路汹之電路 圖。 & 1-321953 VIT05-0273 18607twf.doc/g 圖3為儲存單元21之晝面縮小比為5 :丨的對照矛。 圖4為工業影像規格YC点c r 422之部分像素點^佈 園〇 圖5為工業影像規格YC β c r 420之部分像素點分佈 国。 圖6為本發明另一實施例之畫面縮放電路之電路方塊 I ° 圖7為本發明實施例之晝面縮放方法之流程圖。 【主要元件符號說明】 100、 200 :晝面縮放電路 101、 102、20、60 :緩衝器 111、112、21、61 :計算單元 22、 62 :控制單元 23、 63 :儲存單元 401、402、403、404、5(H、502 :像素點 L1〜L7、L11〜L10 :資料傳輸線In terms of video playback, in order to apply different viewing users and images of different natures, different display image specifications are often used in the same display such as a television. In general, the commonly used display image specifications are Nati〇n Television System Committee (NTSC) and Suntite Aitemati〇n by Line (PAL). However, in both systems, the horizontal scanning frequency and the vertical scanning frequency of the number of scanning lines are different. At the same time, in NTSC and pAL systems, there are different brightness and chroma (chr〇mi face (4) specifications, so the conversion of different image specifications is required to pass through the kneading process circuit. The data transmission lines U to L5 used in the original image are transmitted through the 昼, in order to directly output the data transmission line u, and the voltage of the f _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ^Through ^fH4, it is said that after iG2 storage, and then through the sixty-seven early days and the data transmission = out to the data transmission, then this picture r.. 1 1〇2 and the different early ln, 112 In order to complete the reduction of the screen 5 1-321953 VIT05-0273 18607twf.doc/g can be. From the above, it will be known that this will result in the lack of flexibility in the design of the written circuit. _ 5] 昼 昼 缩放 缩放 缩放 100 100 。 。 。 。 。 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般 一般'(4) plus ~ and an average of five (four) pass round such as the transmission line of the data transmission line = examples V13, V14 and V15, can be lost The voltage value is v\\ + F12 + V13 + VIA + factory 15 5 So - come, the circuit of this design can not only be used and the number of components is too much 'will lead to higher production costs. 1 in ^; number: complex coupling relationship, It will reduce the reliability of the production process. Because of the diversity of the image, and the adjustment according to the image content, the processing circuit of the feather has many components and lacks elasticity. [Invention] [...] The purpose is to provide a kind of face-to-face scaling circuit to process the structure of image data by recursive method, which has less circuit area, lower production cost and improve yield, and second, different image conversion and joining. The new image specifications. '. The purpose of the invention is to provide a face-to-face zoom to the circuit area and production cost, and to improve the yield, and to match the image conversion and add new image specifications. The invention provides a facet zooming circuit, which comprises a buffering crying, an I want to convert the unit and a control unit it. (4) The unit subtracts the original image specification 2 VIT05-0273 18607twf.doc/g Like the relationship between the specifications, to determine the content of the control signal and send a ^ red buffer buffer. Slowly _ receive this control signal to determine whether the image column is stored as a buffer image column, and decide whether to output the stored two-shirt image Column. The different unit of the counting unit to the buffer: output buffer image column, and receive the input image column to perform preset calculation on the column and the buffer image column, and generate the aforementioned storage: the date of the hair Compared with the image conversion circuit described in the embodiment, the required conversion mode is used to change the image format of the control signal, and the control unit (4); The converter between the specifications transmits this control signal to the method to determine the content of the empty signal and remember the body. , 'His. The storage unit includes a non-volatile note. The invention further proposes a recording signal to determine whether or not to perform the method, including: controlling whether the output is stored as a buffered image sequence and determining, in addition, and receiving - the rounding of 1, Receiving from the buffered image column, the image is pre-designed, and the input image column and the buffered image signal are buffered to query the storage unit to issue control and determine the round entry of the material receipt unit. The detour structure that is input to the buffer. 1321953 VIT05-0273 18607twf.doc/g Reduce circuit area, reduce production cost, and increase, add different units to the 'regulation: can cooperate with a variety of the above and other easy to understand the present invention' And with the following as shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Fig. 2 is a block diagram of an embodiment of the present invention. Face scaling circuit 2. . Includes - buffer 2. The second circuit is a control unit 22 and a storage unit 23. The storage star 1 is earlier than 21, and the conversion method is required when the specification is used. This image specification 2 is generated by storing the converted image specifications. The storage unit 23 does not have a == different image can store a plurality of conversion modes to provide to the control mode, 22 output control signals, when the control unit 22 is to be used. When controlling the content of the unit number, it can be referred to as the conversion method of the image specifications required by the Korean exchange party 4 (4), and then the output control signal to the buffer. The device 2 buffer: ς ; whether (4) is coupled to the output of the buffered shadow unit 21 one end is coupled to the slow. Ten 'early 21. The buffered image column is counted and received, and the buffer image sequence is preset for the input image column to store the output image column. In this embodiment ^ can be a non-volatile memory. Cry to cry tn by the control of the remaining system, the wheel out of the shadow system to the slow port 're-output this buffer image column to the calculation unit 21 and the input image 8 1321953 V1T05-0273 18607twf.doc / g column and then do the operation back the way. There may be execution-appropriate weighting. For example, when the value of the transmission line is used, Figure 3 shows the reduction ratio of the storage unit as... The smaller the ratio is 5: 1 The white ij oblique knives are the same as the same day. See Figure 3 and Figure 2, and the comparison table of Figure 3 i, ten, table two rush! §' 20 stores the output image column. 2nd to 4th, ,, or not, is recursive. In each-time recursion, the buffer 20^ is stored before: :::: 3: the input is input to the calculation unit 21 to be like the new wheel i = 裘_ 20 Re-storing the calculated output = moving time ' indicates that the image column will be output first to ft 21 to perform the fourth calculation action, to the output image column of the early #. At this time, the calculation unit outputs the image value. The relationship between the output signal voltage can be exemplified by the transmission voltage V2, V22, V23, V24, and the 5th order data transmission K21 + V22 - factory 23 1-+ V24 2 +V25, the wheel voltage can be 1/ 01 丄]/00 2 In this recursive mode, and the conversion between the image specifications of the storage unit 23 is not matched. At the same time, the storage unit 5 can be compared with the table to obtain the updated optimal comparison table after the test, and the image can be stored lightly. The screen zooming unit 200 only uses the buffer and the new shirt image specifications. Long-term state 2〇 and 呌 OD __ The calculation of the recursive method can save circuit components and surface uT, and 21 units can be used in vertical and horizontal directions.缩放 Zoom < Convert. Figure 4 shows the distribution of partial pixels of the industrial image specification YC Wan Cr 7422. 9 is called 1953 VIT05-0273 18607twf.d〇c/g. Figure 5 is a partial distribution of pixel points of the industrial image specification YC/5Ct 420. Please refer to FIG. 4 and FIG. 5 at the same time, the pixel points 401, 403 and 501 contain brightness data and chroma data (the brightness data is represented by an open circle symbol, the chroma data is represented by a dot symbol), and the pixel points 4〇2, 404 and 502 Contains chroma data. The brightness data transmission voltage of the pixel 501 can be calculated by the picture scaling unit 200 of FIG. 2 to calculate the brightness of the pixel point 4 (the brightness data transmission voltage of Π, 402, 403, 404, for example, the pixel points 4〇1, 402, 403, 404). The data transmission voltage is V401, V402, V403, V404 (V403, V404 quantity K40U F402 ~2 笮氕f ί), then the luminance data transmission voltage of pixel 501 is + V403 2 + V404 :: the reason is no brightness data, so the brightness The data transmission voltage value depends on 5 I turn: the other pixels of square 2 5 can be deduced by analogy. However, the conversion mode of FIG. 4 to FIG. 5 is not limited to the pixel of the four-pixel return number. The conversion of any of Figure 5 can be performed via a facet scaling plane. The difference from Figure 4 to the image data is calculated in a recursive manner to compensate for Figure 6 which is a full picture of another embodiment of the present invention. ^面(四) Wei's circuit block scaling circuit converts the image, and only makes 2^n~L!5' through the control of the face: and the storage unit 63 that needs to be converted when the calculation unit is changed to the image specification is stored and transferred* The image specification is ϋ应于两不10 VIT05-0273 18607twf.d The oc/g is generated with the image specification. The control unit A 62 outputs the control signal. When the control 22 determines the content of the output control signal, the storage unit 63 f can first query the conversion mode between the image specifications. According to the conversion mode, the content of the control signal is determined. The control unit outputs the control signal to the buffer of 60', and the buffer 6〇 can select whether to rotate the image frame according to the control signal, and whether the mosquito is output or not. Buffering image. Ten early 7L01. One end of the computing unit 61 is coupled to the buffering boundary 2〇, the buffered image sequence is output, and receives the input image sequence to input the image column and the buffer image column of the data transmission lines L11 to L15. In the embodiment, the storage unit 63 can be a non-volatile memory. In this embodiment, the circuit is provided with a storage unit 63, a control unit 62, a buffer 60, and The function of r Μ 昼 昼 昼 昼 。 。 。 。 。 。 。 。 。 。 。 。 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平 水平The reduction is taken as an example, but it is known that the technique can be used in the same way, and the difference is in the number of transmission lines of the input/output of the respective units (10). Figure 7 is an embodiment of the present invention. The method of scaling the surface scaling method comprises the following steps: First, in step s7〇i, determining whether to store an output image column as a buffer image according to the - signal signal. Stored as a buffer image column to determine whether to rotate the stored buffer image column. In step S7G3, receive the buffer image column receive-input image column to perform two pre-designs on the input image column and the buffer image column, and generate - Output image column. In step S7〇5, the health of 1-321953 VIT05-0273 18607twf.doc/g is a conversion method required for converting image specifications, which is caused by the relationship between two different image specifications. For the content, first query the conversion between the required image specifications. ^ Control signal. The conversion control signal (4), ^, and then according to the original image specification and the step S707 to be converted into the control signal Content. This 4-sided zoom 1 relationship, non-volatile memory is stored in the converted image specification. In the above-described conversion square wire, in the present invention, the scaling circuit and the complex use control unit provide control signals to the buffer, which are buffered~, due to the structure, the circuit components and the area are greatly reduced: and IS , production costs and improve the reliability of the production process. And with the storage of two balls: a variety of comparison tables' to increase the flexibility of processing image specifications and achieve: two or two performance. Due to the adoption of recursive and achieving better turnarounds, reducing the circuit area, reducing production costs and increasing yields, the company has different image conversions and added new image specifications. And, who is familiar with this skill, can not make a difference with the spirit of the invention ε * can make some changes and retouching, because the definition of the application for the patent garden is the protection of J (simple figure) Explanation] 3 is the circuit diagram of the 缩放 plane scaling circuit 1 。. 2 is a circuit diagram of a picture scaling circuit according to an embodiment of the present invention. & 1-321953 VIT05-0273 18607twf.doc/g Figure 3 is a control spear of the storage unit 21 with a reduction ratio of 5: 丨. Fig. 4 is a partial pixel point of the industrial image specification YC point c r 422. Fig. 5 is a partial distribution of pixel points of the industrial image specification YC β c r 420. FIG. 6 is a circuit block diagram of a picture scaling circuit according to another embodiment of the present invention. FIG. 7 is a flowchart of a method for zooming in a plane according to an embodiment of the present invention. [Main component symbol description] 100, 200: facet scaling circuits 101, 102, 20, 60: buffers 111, 112, 21, 61: computing units 22, 62: control units 23, 63: storage units 401, 402, 403, 404, 5 (H, 502: pixel points L1 to L7, L11 to L10: data transmission line
1313