200945879 九、發明說明： 【發明所屬之技術領域】 【先前技術】200945879 IX. Description of the invention: [Technical field to which the invention pertains] [Prior Art]
求象獲得更好的顯示效果或者有特殊的顯示要 常常需要對圖像進行縮放處理。這裏所㈣s像不僅 ::靜態圖像’也包括動態圖像’如影片圖像。在影片圖 像中’根_描格式的不同，還可能包括訊框格式（fr_ ^at)圖像和場格式圖像（制f〇mat)。特別對於影片 圖像，在至少以下幾種情況下—般需要進行縮放處理… 是片源格式的轉換。由於輸人片源可能有多種長寬比 (aspect rati。），另-方面，影片後處理系統的輸出可能需 要同時支援多種長寬比，例如4:3和16:9的標準畫質電視 (SDTV )’這需要影片後處理系統能夠根據要求將輸入片 源調整到合適的解析度上。二是子母畫面以及小畫面的顯 不子母畫面應用中，除了正常的全螢幕畫面，還需要將 輪入影片縮小到小視窗顯示。三是視窗大小的調整。在某 些應用中，甚至有可能需要支援對視窗作任意比例縮放。 對圖像進行縮放處理’是把原始圖像對應地放大或縮 小到與原始圖像長寬比不相同的目標圖像。圖像縮放處理 常常採用濾波器’通過濾波器而達到縮小或放大的目的。 濾波器通常針對圖像中的圖元來進行濾波。一般來說，濾 波器從源圖像中取多個畫素點插值滤波後形成一個目標圖 6 200945879 像的晝素點，由此類推，得到所有目標圖像的畫素點。每 個晝素點的濾波常常會涉及多個縮放參數，例如濾波函 數、濾波器階數、插值量化精度以及放縮比例等。採用不 同的縮放參數進行插值所獲得的目標圖像是不相同的。現 有技術中，通常對圖像縮放裝置一次性配置濾波所用的縮 放參數，並以固定的參數處理相關的縮放操作。如果圖像 特徵發生變化，或者顯示需求發生變化，例如顯示螢幕的 長寬比從4 · 3變化為16 : 9 ’則圖像縮放裝置的軟硬體結 〇 構都需要進行較大的改動。 有一些改進的技術方案中，還可能在對每一圖像進行 處理前，it行一次圖像縮放裝置的參數的配置，以滿足對 不同幅有不同參數的縮放處理。但在多視窗的圖像縮放 時’每個視窗可能有不同的縮放要求，需要配置不同的縮 放參數’甚至同一個視窗也有不同參數的縮放要求，例如 在非線性拉伸操作等情況下。通常縮放操作的參數是由外 部中斷後’以軟體配置的。而每次配置不同的縮放參數都 9 冑要等待中斷，造成大量延遲。這樣勢必導致縮放處理逮 度大幅降低，特別在即時影片播放時’會給後面的播放帶 來比較大的影響。 目前提出的另-種改進方案是把圖像分成多個區塊。 對於每塊圖像，預先配置㈣_像縮㈣作所需的參 數’在預先設置的固定區域内縮放參數都是固定的。例如， 第1圖所示的圖像分成7個區塊，由於每區塊圖像的 縮放操作所需的參數都要預先設置並存放在緩衝設 這給圖像縮放裝置整體的存儲帶來了負擔。並且，由於圖 200945879 像是按照固定的區域分塊，因此，無法按照需求靈活地配 ' 置圖像的非線性縮放範圍。 本質上歸納起來，目標圖像上的每一個晝素點，都是 利用原始圖像上某一些有一定關聯的畫素點，根據事先確 定的一些參數，通過一定的計算步驟而得到的，因此如果 能约處理以上所有的變化因素，確切的說，需要能夠精確 的獨立控制目標營幕上每一個畫素點相關的所有的以上計 算參數，就有可能在一個統一的構架下面進行任意的圖像 參 縮放。當然這個代價無疑是非常巨大的，標準畫質模式下 就存在720*576，共計414720個畫素點，高晝質模式下更 是存在1920*1080 ，共計20736〇〇個畫素點。採用硬體存 儲的話面積將過大而無法接受。而採用軟體處理方式的 話，由於逐個畫素點都需要獨立配置，延遲和計算的成本 CPU也無法承受。 因此，業界正在努力尋求一種在圖像特徵或顯示需求 等發生變化時，能靈活配置圖像的非線性縮放範圍，甚至 ❿ ㊣對多個視窗的圖像分別進行非線性縮放又能消耗較少存 儲面積的圖像縮放裝置。 【發明内容】 為克服現有技術中存在的缺陷，本發明提出了一種通 用的圖像縮放裝置’可按需求任意給不同視窗配置不同的 縮放參數’也可給視窗内的不同位置配置不同的縮放參 數’並能降低存儲需求。 本發明並提供了一種相應的圖像縮放方法，該方法可 200945879 需长任意，，’Q不同視窗配置不同的縮放參數，也可給視窗 内的不同位置配置不同的縮放參數，並能降低存儲需求。 根據本發明的一個方面，提供一種通用圖像縮放裝 置，包括： 一圖像縮放單元，用以對一圖像進行一垂直縮放及一 水平縮放； 一控制單元，用以接受外部輸入之至少一圖像縮放參 數並對圖像縮放單元之圖像縮放操作進行控制； ❹ 一計數單元，用以記錄對應圖像縮放參數之一預定使 用量並在圖像縮放參數的預定使用量完成時，控制單元 更新圖像縮放參數。 上述通用圖像縮放裝置中，其中控制單元包括存放該 圖像縮放參數之至少一圖像縮放參數暫存器。 上述通用圖像縮放裝置中，其中計數單元包含對圖像 縮放參數之預定使用量進行計數之至少一計數器。 上述通用圊像縮放裝置中，其中該控制單元包括複數 ❹ 個圖像縮放參數暫存器，計數單元包括與等圖像縮放參數 暫存器相對應的複數個計數器。 上述通用圖像縮放裝置中，其中計數單元包括一垂直 方向計數器，係用以根據使用相同之圖像縮放參數之至少 一水平行之數量，記錄圖像縮放之垂直方向之預定使用量 並進行計數，並在所述垂直方向預定使用量到達時，由計 數單元提示控制單元更新圖像縮放參數及對應垂直方向之 預定使用量。 上述通用圖像縮放裝置中，其中垂直方向計數器為全 200945879 域計數器。 ^述通用圖像縮放裝置更包含—預測配置單元，預測 置單70係用以檢剛圖像之圖像縮放參數，並根據圓像縮 放參數預測下1像的之®像縮放參數。 f述通用圖像縮放裝置中，其中圖像縮放參數包括滤 ’函插值3：化階數、步進值、濾波器階數或其組合。 上述通用圖像縮放裝置中，其中預定使用量係為圓像To obtain a better display or to have a special display, it is often necessary to scale the image. Here, the (4)s image not only includes: still images 'including dynamic images' such as film images. In the movie image, the 'root_scanning format' may also include a frame format (fr_^at) image and a field format image (f〇mat). Especially for movie images, zooming is required in at least the following cases... It is a conversion of the source format. Since the input source may have multiple aspect ratios, on the other hand, the output of the film post-processing system may need to support multiple aspect ratios, such as 4:3 and 16:9 standard picture TVs ( SDTV) 'This requires the film post-processing system to adjust the input source to the appropriate resolution as required. Second, in the picture of the mother-in-law and the small-screen display, in addition to the normal full screen, it is also necessary to reduce the rounded video to the small window display. The third is the adjustment of the window size. In some applications, it may even be necessary to support any scaling of the window. The scaling process of the image is to enlarge or reduce the original image correspondingly to a target image that is different from the original image aspect ratio. Image scaling processing often uses a filter' to achieve reduction or amplification through a filter. Filters are typically filtered for primitives in the image. In general, the filter extracts multiple pixel points from the source image to form a pixel point of the target image, and so on, to obtain the pixel points of all the target images. The filtering of each pixel point often involves multiple scaling parameters, such as filtering functions, filter order, interpolation quantization precision, and scaling ratio. The target images obtained by interpolation using different scaling parameters are not the same. In the prior art, the scaling parameters used for filtering are typically configured once for the image scaling device, and the associated scaling operations are processed with fixed parameters. If the image characteristics change, or the display requirements change, for example, the aspect ratio of the display screen changes from 4 · 3 to 16: 9 ', the soft and hard structure of the image scaling device needs to be changed. In some improved technical solutions, it is also possible to configure the parameters of the image scaling device once before processing each image to satisfy the scaling process with different parameters for different amplitudes. However, when multi-window image scaling, 'each window may have different scaling requirements, different scaling parameters need to be configured' or even the same window has scaling requirements for different parameters, such as in the case of nonlinear stretching operations. Usually the parameters of the scaling operation are configured by software after being externally interrupted. Each time you configure a different scaling parameter, you have to wait for an interrupt, causing a lot of delay. This will inevitably lead to a significant reduction in zoom processing, especially in the case of instant video playback, which will have a greater impact on subsequent playback. Another improvement proposed at present is to divide the image into multiple blocks. For each image, the pre-configured (four)_images (four) are required for the parameters. The scaling parameters are fixed in the preset fixed area. For example, the image shown in Fig. 1 is divided into 7 blocks, and the parameters required for the scaling operation of each block image are preset and stored in the buffer setting, which brings the storage of the image scaling device as a whole. burden. Moreover, since Figure 200945879 is like a fixed area, it is not possible to flexibly match the non-linear scaling range of the image as needed. In essence, each pixel point on the target image is obtained by using certain pixel points on the original image that are related to each other, according to some parameters determined in advance, through certain calculation steps. If all of the above variables can be dealt with, specifically, it is necessary to be able to accurately and independently control all the above calculation parameters related to each pixel point on the target camp, it is possible to perform arbitrary diagrams under a unified framework. Like the parameter zoom. Of course, this cost is undoubtedly very large. There are 720*576 in the standard image quality mode, a total of 414,720 pixel points, and there are 1920*1080 in the high quality mode, totaling 20,736 pixel points. With hardware storage, the area will be too large to be acceptable. With the software processing method, since the pixel points need to be configured independently, the CPU of the delay and calculation cannot be tolerated. Therefore, the industry is striving to find a non-linear scaling range that can flexibly configure images when image features or display requirements change, and even reduce the nonlinear scaling of images in multiple windows. Image scaling device for storage area. SUMMARY OF THE INVENTION In order to overcome the deficiencies in the prior art, the present invention proposes a universal image scaling device that can arbitrarily assign different scaling parameters to different windows as needed. It can also configure different scaling for different positions in the window. The parameter 'can also reduce storage requirements. The invention also provides a corresponding image scaling method, which can be long and arbitrary, and the 'Q different windows can be configured with different scaling parameters, and different scaling parameters can be configured for different positions in the window, and the storage can be reduced. demand. According to an aspect of the present invention, a general image scaling apparatus includes: an image scaling unit for performing a vertical scaling and a horizontal scaling of an image; and a control unit for accepting at least one of the external inputs The image scaling parameter controls the image scaling operation of the image scaling unit; ❹ a counting unit for recording a predetermined usage amount of the corresponding image scaling parameter and controlling when the predetermined usage amount of the image scaling parameter is completed The unit updates the image scaling parameters. In the above general image scaling apparatus, wherein the control unit includes at least one image scaling parameter register for storing the image scaling parameter. In the above general image scaling apparatus, wherein the counting unit includes at least one counter that counts a predetermined usage amount of the image scaling parameter. In the above general image zooming apparatus, the control unit includes a plurality of image scaling parameter registers, and the counting unit includes a plurality of counters corresponding to the equal image scaling parameter registers. In the above general image zooming apparatus, wherein the counting unit includes a vertical direction counter for recording a predetermined usage amount of the vertical direction of the image scaling and counting according to the number of at least one horizontal line using the same image scaling parameter. And when the predetermined usage amount reaches in the vertical direction, the counting unit prompts the control unit to update the image scaling parameter and the predetermined usage amount corresponding to the vertical direction. In the above general image zooming apparatus, the vertical direction counter is a full 200945879 domain counter. The general image scaling device further includes a prediction configuration unit for predicting image scaling parameters of the image and predicting the image scaling parameter of the next image based on the circular image scaling parameter. In the general image scaling apparatus, the image scaling parameter includes a filter terminology 3: a quantization order, a step value, a filter order, or a combination thereof. In the above general image zooming device, wherein the predetermined usage amount is a circular image
縮放參數之使用·欠數、使用比例、使用時間或使用長度。 根據本發明的另一方面，提供一種通用圖像縮放方 法，包括以下步驟： 配置至/ —組圖像縮放參數及與該組圖像縮放參數 相對應的預定使用量； b對該圖像上之等晝素點進行圖像縮放處理，並對該 組圖，縮放參數已經完成之複數個使用量進行計數； —w該、，'且圖像縮放參數已經完成之使用量達到配置的預 疋使用量時，重新配置下一組圖像縮放參數及與該組圖像 Φ 縮放參數相對應的預定使用量。 上述通用圖像縮放方法還包括步驟： d在备别的一幅圖像縮放操作完成後，根據該幅圖像 之圖像縮放參數關下—幅圖像之-圖像縮放參數。 上述通用圖像縮放方法中，其中圖像縮放參數包括渡 波器之抽頭數、放縮比例、據波函數、插值量化階數或其 組合。 工地迎用團像縮放方法中， 一放縮比例 維圖像之一抽頭數 濾波函數、一插仓 200945879 化階數或其纟且合。 ^通用圖像縮放方法中，其中圖像縮放參數之 使用量係為該圖像縮放參數之使用次數、使用比例、使 時間或使用長度。 上述通用圖像縮放方法中，其中圖像縮放參數係以一 目標圖像為物件之參數。Use of scaling parameters, undercount, usage ratio, usage time, or length of use. According to another aspect of the present invention, there is provided a general image scaling method comprising the steps of: configuring a /-group image scaling parameter and a predetermined usage amount corresponding to the group of image scaling parameters; b on the image The pixel points are subjected to image scaling processing, and the number of usages of the set of graphs and the scaling parameters have been counted; -w,, and the image scaling parameter has been completed and the usage amount reaches the configured preview. When the amount is used, the next set of image scaling parameters and the predetermined usage amount corresponding to the set of image Φ scaling parameters are reconfigured. The above general image scaling method further comprises the steps of: d after the completion of the image scaling operation of the image, the image scaling parameter of the image is turned off according to the image scaling parameter of the image. In the above general image scaling method, the image scaling parameter includes the number of taps of the ferrator, the scaling ratio, the wave function, the interpolation quantization order, or a combination thereof. In the construction site, the group image scaling method is used. One of the scales of the scale image is one of the number of taps. The filter function, a plug-in position, 200945879, or the order. In the general image scaling method, the usage amount of the image scaling parameter is the number of uses of the image scaling parameter, the usage ratio, the time of use, or the length of use. In the above general image scaling method, the image scaling parameter takes a target image as a parameter of the object.
上述通用圖像縮放方法中，其中步驟a更包含一步驟: 配置根據使用相同之圖像縮放參數之至少—水平行之數量 所確疋之該圖像縮放之一垂直方向之該預定使用量。 上述通用圖像縮放方法中，其中步驟a更包含一步驟： 預先配置一圖像的所有圖像縮放參數的步驟。 上述通用圖像縮放方法中，其中步驟b中對該圖像縮 放參數已經完成之該使用量進行計數的步驟更包含用自減 或自加的方式進行計數之步驟。 本發明利用計數單元對每組縮放參數的使用次數進行 獨立計數，因此能夠獨立控制目標螢幕上任意圖元點的所 有縮放參數，從而可以在一個統一的通用縮放結構下支援 多視窗的圖像縮放，並可支援一維或多維圖像縮放演算法 或者演算法組合。由於對縮放參數採用計數方式，相同的 縮放參數不必重複進行存放，因而可以節約大量存儲空間。 在參閱圖式及隨後描述之實施方式後，該技術領域具 有通常知識者便可瞭解本發明之目的，以及本發明之技術 手段及實施態樣。 【實施方式】 200945879 第2A圖為根據本發明的一種通用圖像縮放裝置之方 塊圖。參見第2A圖，本發明的圖像縮放裝置包括圖像縮放 單元21和對圖像縮放草元的處理過程進行控制的控制單元 22。圖像縮放單元可對圓像進行垂直縮放和水準縮放。控 制單元接受並存儲外部輸入的縮放參數，對圖像縮放單元 的縮放操作進行控制。本發明在圖像縮放裝置中並設置了 計數單元23，計數單元能夠記錄縮放參數的預定使用量， 如使用次數等，並在縮放參數已經完成的使用量到達預定 使用量時’提示控制單元更新縮放參數。 圖像縮放單元中通常利用濾波器對原始圖像進行插值 來獲得目標圖像。圖像縮放的主要方法是根據放縮比例， 即目標圖像長度與原始圖像長度的比值，確定出目標圖像 中需要插值的點的位置，然後確定插值點位置所對應的原 始圖像的多個原始點，這些原始點經過插值濾波器，根據 係數加權計算得到所需要插值的目標樣點。不同的插值濾 波器可以得到不同的插值圖像效果。 第3圖係為根據本發明的一個實施例的據波器之一示 意圖。濾波器所抽取的資料樣點也稱為取樣的階數（tap)。 P白數了以根據顯示精確度的要求來設置。於本實施例中， 第3圖中濾波器示具有4階數。理論上，對於每個圖像畫 素點的縮放處理，可以取不同的顯示精確度值，即階數。 遽波器從源資料巾抽取多個原始點數據，每個畫素點都乘 上一個與之對應的係數，得到一個乘積，再把這些乘積累 加起來，輸出所得的結果，則可獲得目標點數據。每個目 標點可以在圖像的源資料中選取一個參考點，一般參考點 12 200945879 選在目標點在圖像中的位置的附近。根據放縮比例可以確 定相鄰目標點對應的參考點之間的距離（以點數為單位）， 即步進值（step)。運算中對於每一個目標點，只需要知道 其前-點的參考點和前一點的步進值，就可以根據該點的 步進值得到該點的參考點，以此類推，從而得到遽波所需 的全部原始點。圖像的放縮比例既可能是固定的絕對放縮 比例，也可能是非線性縮放中一個隨位置變化的放縮比 例。因此，放縮比例可以根據實際的縮放需求來變化，理 ® 論上甚至可以逐點變化。 第4圖為根據本發明一實施例的濾波器的插值原理示 意圖。圖中，標號41表示目標圖像，標號411表示目標晝 素點4U ’標號42表示原圖像，標號42卜422、423和424 分別表示原晝素點i至原晝素點4，標號43表示濾波器視 窗，橫向軸表示目標點到X的距離，縱向軸表示目標點的 權重’標號45表示濾波函數f(x)，標號s表示步進值。 根據需要的圖像放縮比例，可以確定目標圖像中每兩個插 ❹ 值點相對於原圖像中的步進值。根據插值點的位置，可以 得到原圖像中參與插值的N個點，原圈像中的n個點與插 值濾波器所計算出來的加權值作加權和就得到了目標圖像 中插值點的畫素。從以上過程可以看出，對於每一個插值 點，首先需要確定插值距離χ，然後根據所使用的濾波函數 逐點计算出目標點的加權值’即插值係數，然後才能進行 加權’而得到插值點。不同的濾波函數的表現是不相同的， 它們引入的誤差也是不同的。對於每個圖像樣點的縮放處 理’可以選取不同的濾波函數。通常需要根據具體應用來 13 200945879 選擇適當的濾波函數。 在實際硬體實現中，採用逐點計算插值係數的方法代 價太大，通常所採取的做法是：設定一定的精度’也稱為 量化階數。根據量化階數對插值距離X做量化，例如，量 化階數為8，則對插值距離X作8段量化，將X值量化成： 0, 1/8, 2/8, 3/8, 4/8, 5/8, 6/8, 7/8。根據這些量化後的 X 值預 先計算出濾波函數所對應的插值係數，可以獲得8組插值 係數。然後，將該8組插值係數以表格的形式存放於記憶 ❹ ❹ 體中°在插值過程中，根據每個晝素點得到的插值距離x， 選取最接近的插值係數做加權計算。為了得到更高的精 度’也可以把量化階數設為16段’ 32段或更高的精度。插 值係數也稱為濾波係數，一般地，在設定了濾波函數和量 化階數後，則可以得到確定的濾波係數的索引表。 對於圖像中的每個畫素點來說，確定了縮放參數，即 可得到確疋的縮放結果。例如，根據前文所述，一維圖像 的縮放操作中，縮放參數包括遽波函數、量化階數、放縮 比例抽頭數。確定了上述縮放參數，原始圖像縮放後可 得到唯一確定的目標圖像。 第5圖為根據本發明_個實施例的通用圖像縮放裝置 I 4數單7C與控制單元的縮放參數暫存器的對應關係示 =參見第5圖，控制單元22中設有至少一組縮放參數 !^ 225’用來存放從外部輪人的縮放參數。可以預先確 ϋ Ϊ =震置在進行縮放操作時可能用到的縮放參數 = 放參數個數來設定-組縮放參數暫存器的 個數每:人料部輸人㈣錢時，㈣輸人浦放參數 200945879In the above general image scaling method, the step a further comprises a step of: configuring the predetermined usage amount of one of the vertical directions of the image scaling determined according to at least the number of horizontal lines using the same image scaling parameter. In the above general image scaling method, step a further includes a step of: pre-configuring all image scaling parameters of an image. In the above general image scaling method, the step of counting the usage amount in which the image scaling parameter has been completed in the step b further includes the step of counting by self-decrementing or self-adding. The invention utilizes the counting unit to independently count the number of times of using each set of scaling parameters, so that all the scaling parameters of any pixel point on the target screen can be independently controlled, thereby supporting multi-window image scaling under a unified universal zoom structure. And can support one-dimensional or multi-dimensional image scaling algorithms or algorithm combinations. Since the scaling parameters are used for counting, the same scaling parameters do not have to be stored repeatedly, which saves a lot of storage space. The object of the present invention, as well as the technical means and embodiments of the present invention, will be apparent to those skilled in the art in the light of the appended claims. [Embodiment] 200945879 Fig. 2A is a block diagram of a general-purpose image scaling device according to the present invention. Referring to Fig. 2A, the image scaling apparatus of the present invention includes an image scaling unit 21 and a control unit 22 that controls the processing of the image scaling curs. The image scaling unit can vertically scale and level the circle image. The control unit accepts and stores the scaling parameters of the external input to control the scaling operation of the image scaling unit. The present invention is provided in the image scaling apparatus and is provided with a counting unit 23 capable of recording a predetermined usage amount of the scaling parameter, such as the number of uses, etc., and prompting the control unit to update when the usage amount of the zooming parameter has reached the predetermined usage amount. Scale the parameters. In the image scaling unit, the original image is usually interpolated using a filter to obtain a target image. The main method of image scaling is to determine the position of the point in the target image that needs to be interpolated according to the ratio of the scaling, that is, the ratio of the length of the target image to the length of the original image, and then determine the original image corresponding to the position of the interpolation point. A plurality of original points, which are subjected to an interpolation filter, and the target samples to be interpolated are obtained according to coefficient weighting. Different interpolation filters can get different interpolation images. Figure 3 is a schematic illustration of one of the wavers in accordance with one embodiment of the present invention. The data sample extracted by the filter is also referred to as the sampling order (tap). P is whited to be set according to the requirements of display accuracy. In the present embodiment, the filter shown in Fig. 3 has a fourth order. In theory, for the scaling process of each image pixel point, different display accuracy values, ie, orders, can be taken. The chopper extracts a plurality of original point data from the source data towel, and each pixel point is multiplied by a corresponding coefficient to obtain a product, and then the multiplication and accumulation are added up, and the obtained result is output, and the target point is obtained. data. Each target point can select a reference point in the source data of the image. The general reference point 12 200945879 is selected near the position of the target point in the image. According to the scaling ratio, the distance (in points) between the reference points corresponding to adjacent target points can be determined, that is, the step value. For each target point in the operation, only need to know the reference point of the front-point and the step value of the previous point, you can get the reference point of the point according to the step value of the point, and so on, so as to obtain the chopping All the original points needed. The scaling ratio of an image may be either a fixed absolute scaling ratio or a scaling ratio change with position in nonlinear scaling. Therefore, the scaling ratio can be changed according to the actual scaling requirements, and the theory can be changed point by point. Fig. 4 is a diagram showing the principle of interpolation of a filter according to an embodiment of the present invention. In the figure, reference numeral 41 denotes a target image, reference numeral 411 denotes a target pixel point 4U', reference numeral 42 denotes an original image, and reference numerals 42 422, 423 and 424 denote a primary pixel point i to an original pixel point 4, respectively. The filter window is shown, the horizontal axis represents the distance from the target point to X, and the vertical axis represents the weight of the target point. The reference numeral 45 represents the filter function f(x), and the reference numeral s represents the step value. According to the required image scaling ratio, it is possible to determine the step value of each of the two interpolation value points in the target image with respect to the original image. According to the position of the interpolation point, N points participating in the interpolation in the original image can be obtained, and the n points in the original circle image are weighted with the weighted values calculated by the interpolation filter to obtain the interpolation point in the target image. Picture. It can be seen from the above process that for each interpolation point, the interpolation distance χ needs to be determined first, and then the weighted value of the target point, ie, the interpolation coefficient, is calculated point by point according to the filter function used, and then the weighting can be performed to obtain the interpolation point. . The performance of different filter functions is different, and the errors introduced by them are also different. Different filtering functions can be selected for the scaling process of each image sample. It is usually necessary to select the appropriate filter function according to the specific application 13 200945879. In the actual hardware implementation, the method of calculating the interpolation coefficient point by point is too expensive, and the usual approach is to set a certain precision 'also called the quantization order. The interpolation distance X is quantized according to the quantization order. For example, if the quantization order is 8, the interpolation distance X is subjected to 8-segment quantization, and the X value is quantized into: 0, 1/8, 2/8, 3/8, 4 /8, 5/8, 6/8, 7/8. According to these quantized X values, the interpolation coefficients corresponding to the filter function are calculated in advance, and eight sets of interpolation coefficients can be obtained. Then, the 8 sets of interpolation coefficients are stored in the form of a table in the memory ° °. In the interpolation process, the nearest interpolation coefficient is selected according to the interpolation distance x obtained for each pixel point for weighting calculation. In order to obtain higher accuracy, the quantization order can also be set to an accuracy of 16 segments '32 segments or higher. The interpolation coefficient is also called a filter coefficient. Generally, after the filter function and the quantization order are set, an index table of the determined filter coefficients can be obtained. For each pixel point in the image, the scaling parameter is determined, which results in a definite scaling result. For example, according to the foregoing, in the scaling operation of the one-dimensional image, the scaling parameters include a chopping function, a quantization step, and a scaling ratio tap number. The above scaling parameters are determined, and the original image is scaled to obtain a uniquely determined target image. Figure 5 is a diagram showing the correspondence relationship between the general image scaling device I 4 and the scaling parameter register of the control unit according to the embodiment of the present invention. Referring to Figure 5, at least one group is provided in the control unit 22. The scaling parameter!^ 225' is used to store the scaling parameters from the outside wheel. It can be pre-determined Ϊ 震 震 震 震 震 震 震 震 震 进行 震 进行 震 震 震 震 震 震 震 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放 缩放Pufang parameter 200945879
所需進行使用的預定使用量。計數單元23設置了計數器 235，能夠記錄縮放參數的預定使用量，並對縮放參數已經 完成的使用量進行計數’在縮放參數的預定使用量完成 時，由控制單元更新縮放參數。例如，把縮放參數的預定 使用量設定為計數器的計數初始值，縮放參數每使用一次 後《十數器進行自減，當計數器自減至零值時，提示控制 單元更新縮放參數暫存器中的縮放參數。因此，可以根據 實際需求預先配置料備進行圖像·處理的縮放參數， 在進行到適㈣時機時，再錢魏參數，並制更新後 的縮放參數繼續進行圖像縮放處理，不必同時存放大量相 同的縮放參數，因此節約了存儲空間。 根據本發明的一些實施例’計數單元為全域計數器。 全域計數器記錄的多個計數值可轉目標圖像中的一個水 平行，或者-幅目標圖像’劃分成任意寬度的小塊，甚至 可以以比較小的代價獲得對圖像縮放的逐點控制。 縮放參數的預定使用量是根據目標圖像上的圖元來計 算’而非原始圖像上的圖^在—般情況下，由於我們的 效果評估都是針對目標圖像而言的，所有縮放參數變化， 例如影響濾波效果與放縮比例㈣波函數、量化階數、放 縮比例、抽頭數等縮放參數，也都是針對目標圖像而確定 的，這樣做同時也可以減少計算上的_ 1定使用量可 梦自由。X置為使用次數、使用比例、使用時間或使用長度 最終㈣㈣或硬料算統―折換錢用次數後作 為计數器的計數初始值。 根據本發明的一 些實施例，把這些具艎的縮放參數插 15 200945879 父給硬體處理’而由軟體進行目標圖像逐個畫素點 、/配置’則可以用~種通用的硬體結構處理各類圖像 縮放=法’由此可以達到理論上的“逐點控制，，的要求。 配置=明的另一實施例中，圖像縮放裝置並設有預測 數 間 圖 放參數，1根t見第2Β圖)’可以檢測一幅圖像的全部縮 :參：並根據檢測到的資料預測下一幅圖像的 數。^_置單元可崎—步減少裝載縮放參數所需的時 因而提鬲圖像縮放裝置的處理速度。The amount of scheduled usage required for use. The counting unit 23 sets a counter 235 capable of recording a predetermined usage amount of the scaling parameter and counting the usage amount that the scaling parameter has been completed. When the predetermined usage amount of the scaling parameter is completed, the scaling parameter is updated by the control unit. For example, the predetermined usage amount of the scaling parameter is set as the initial value of the counter, and the scaling parameter is self-decreasing after each use of the scaling parameter. When the counter is self-decreased to zero value, the control unit is prompted to update the scaling parameter register. The scaling parameters. Therefore, the scaling parameters of the image processing can be pre-configured according to actual needs. When the timing is appropriate, the parameters are further adjusted, and the updated scaling parameters are continued to perform image scaling processing, without having to store a large amount of images at the same time. The same scaling parameters, thus saving storage space. According to some embodiments of the invention, the counting unit is a global counter. The plurality of count values recorded by the global counter can be transferred to one horizontal line in the target image, or the - target image can be divided into small blocks of arbitrary width, and the point-by-point control of image scaling can be obtained even at a relatively small cost. . The predetermined usage of the scaling parameter is calculated based on the primitive on the target image instead of the image on the original image. In general, since our effect evaluation is for the target image, all scaling Parameter changes, such as scaling factors affecting the filtering effect and scaling ratio (4) wave function, quantization order, scaling ratio, number of taps, etc., are also determined for the target image, which can also reduce the computational _ 1 can be used freely. X is set to the number of uses, the proportion of use, the time of use, or the length of use. Finally (4) (4) or hard material calculation - the initial count value of the counter after the number of times the money is used. According to some embodiments of the present invention, these specific scaling parameters are inserted into the 15200945879 parent to the hardware processing ' while the target image is pixel-by-pixel point, /configuration' can be processed by the general hardware structure. Various types of image scaling = method 'This can achieve the theoretical "point-by-point control," requirements. In another embodiment of configuration = Ming, the image scaling device is provided with predictive inter-picture parameters, 1 t see the second diagram) 'You can detect the total shrinkage of an image: parameters: and predict the number of the next image based on the detected data. ^_ Set the unit to reduce the time required to load the scaling parameters The processing speed of the image scaling device is thus improved.
第6圖為本發明的通用圖像縮放方法的流程圖。第6 該方法包含以下步驟： S1 ’配置至少—組圖像縮放參數和與該組縮放參數相 對應的預定使用量； S2,對圖像上的畫素點進行圖像縮放處理，並對縮放 參數已經完成的使用量進行計數； Μ，判斷縮放參數已經完成的使用量是否達到配置的 預定使用量；如否，程式返回步驟S2;如是，進入步驟§4, 重新配置下一組縮放參數和與該組縮放參數相對應的預定 使用量。 上述圖像縮放方法可以進一步包括如下步驟，在一幅 圖像縮放操作完成後，根據該幅圖像的縮放參數，預測下 一幅圖像的縮放參數。 縮放參數暫存器225和計數器235的數量可以根據需 求來設置，以下以幾個實施例來說明幾種不同的配置。 根據本發明的一個實施例，計數單元包括至少一個計 數器，計數器中記錄一組縮放參數的預定使用量。如第5 200945879 圖所示，縮放參數暫存器225設置了 一組縮放參數的值， 例如包括濾波函數、量化階數、步進值、濾波器階數等， 計數器235中記錄該組縮放參數預定使用量的值。例如， 該組縮放參數需要使用在目標圖像1〇〇個晝素點的縮放處 理上’可以把計數器的初始值設為1〇〇。在每次完成一個晝 素點的縮放處理後，就把計數器的值減丨，這樣，直到完成 100個畫素點的縮放處理’計數器的值變為零，計數單元將 提示縮放參數暫存器裝載一組新的縮放參數，並把一同輸 ❹ 入的該組縮放參數的預定使用量記錄在計數器中。 根據本發明的另一個實施例，計數單元23中也可以設 置一組計數器。計數器的數量與縮放參數暫存器組225中 的縮放參數暫存器的數量相對應，對於每一個縮放參數暫 存器分別配置一個相對應的計數器，可以存放該縮放參數 暫存器中縮放參數的預定使用量。如第7圖所示，第一計 數器231(方框内的文字表示該計數器所進行計數的物件， 下同）與縮放參數暫存器組中的濾波函數暫存器2251對 ® 應，第m 232與縮放參數暫存器組中的量化階數暫 存器2252對應，第三計數器233與縮放參數暫存器組中的 步進值暫存器2253對應，第四計數器234與縮放參數暫存 器組中的濾波器階數暫存器2254對應。 根據本發明的另一個實施例，如第8圖所示，控制單 元中設置了兩組縮放參數暫存器，肖組驗參數暫存器可 以輪流存取❶對應每組縮放參數暫存器，並各設置—個計 數器，在每次刷新-組縮放參數暫存器的縮放參數值時， 同時把該組縮放參數值的預定使用量輪入與其對應的計數 17 200945879 器。具體而言，可以在第一縮放參數暫存器組225—1中存 放當前進行縮放處理需要使用的縮放參數，並在第一計數 器單元23—1中存放第一組縮放參數的預定使用量。在使 用第一組縮放參數進行操作期間，在第二縮放參數暫存器 組225 — 2中存放緊隨其後的縮放操作需要使用的縮放參 數’並把第二組縮放參數預定使用量輸入對應的第二計數 單元23 — 2。在計數單元指示第一組縮放參數暫存器中存放 的當前的縮放參數的預定使用量已經到達，需要變換縮放 ❿ 參數時’直接開始使用第二縮放參數暫存器組中存放的縮 放參數。之後，在使用第二組縮放參數進行操作時，從外 部輸入將要使用的縮放參數刷新第一縮放參數暫存器組， 並把其對應的縮放參數預定使用量輸入第一計數單元。在 計數單元指示第二縮放參數暫存器組中存放的縮放參數的 預定使用量已經到達，需要變換縮放參數時，直接開始重 新取用第一縮放參數暫存器組的值。按此方式將兩組縮放 參數暫存器和對應的兩組計數器輪換使用，可以在利用一 ® 組縮放參數進行操作時，同時進行另一組縮放參數的裝載 工作，加快了處理速度。 根據本發明的另一個實施例，可以設置多組縮放參數 暫存器和多個與其分別對應的計數器。在對目標圖像的某 一行進行縮放處理的時候，可以把所有縮放參數及其對應 的預疋使用量全部記錄下來，同一幅圖像的其他行可以按 照該行的設置進行縮放處理。第9圖為根據本發明該實施 例的縮放參數和縮放參數的預定使用量的分佈原理示意 圖。參見第9圖，假設圖像中的第1行需要進行非線性縮 18 200945879 放’共分成5個部分’每個部分有其對應的縮放參數和縮 放參數預定使用量。例如第9圖中的第一組縮放參數和第 一組縮放參數的預定使用量，第二組縮放參數和第二組縮 放參數的預定使用量’第三組縮放參數和第三組縮放參數 的預定使用量’第四組縮放參數和第四組縮放參數的預定 使用量’第五組縮放參數和第五組縮放參數的預定使用 量。由於計數單元中記錄了縮放參數的預定使用量，並可 根據預定使用量來提示更換縮放參數，因此5個部分之間 ❹ 的界限是可以靈活設置的’每個部分不再被限定成固定的 大小和始末位址’可以對應5個部分分別配置相應的縮放 參數暫存器和計數器。對於其後的行，如果仍然分成相同 的5個部分並分別採用相同的縮放參數時，例如第9圖中 的第2至6行，仍可以直接沿用已經配置好的縮放參數和 縮放參數預定使用量’而不必重新裝載5個部分各自的縮 放參數，因此也提高了處理速度。每一行的縮放參數的劃 分也元全可以按照貫際需求來確定，例如第9圖中的第7 Φ 到10行還可以分成4個部分。可以對應這4個部分分別配 置相應的縮放參數暫存器和計數器。 對於一幅圖像，計數單元還可以在圖像的垂直方向設 置垂直方向計數器，用於記錄一個水平行上的多組縮放參 數在垂直方向的預定使用量，並在垂直方向的預定使用量 到達時，提示更新縮放參數。垂直方向的預定使用量可以 以行數為單位。例如，帛9圖中第1至6行的多組縮放參 數是相同的，在垂直方向計數器中記錄與其相應的垂直方 向預定使用量，在圖像的垂直方向每完成—行縮放㈣， 200945879 广把垂直方向。十數暫存器的值減i。直到垂直方向計數暫存 器的值減為零，計數單元指示垂直方向縮放參數的預定使 用，已經達到了’再開始裝載第7行的縮放參數。這樣， 在當前行與上—行使用相同的縮放參數的情況下，在切換 行，行圖像縮放時可省略重新裝載縮放參數和縮放參數的 預定使用量的動作，因而可以減少中斷次數，並節約裝載 縮放參數的功耗和時間，同時也達到了靈活地對一幅圖像 分成多個不同部分分別按照不同的縮放參數進行縮放處理 φ 的目的。 在本發明的一個實施例中，圖像縮放裝置可以預先配 置一幅圖像的所有縮放參數。先針對水平行設定多組縮放 參數，並在相應的計數器中配置每組縮放參數對應的預定 使用量；再配置多個垂直方向計數器’每個垂直方向計數 暫存器中設置根據同組縮放參數所應用的水平行的數量所 確定的縮放參數垂直方向預定使用量。 在進行多視窗的圖像縮放處理中，對於每個視窗可以 ® 各自設置縮放參數暫存器和相應的計數器。縮放參數暫存 器中可以預先存放各視窗的縮放參數，計數單元分別對每 個視窗縮放參數的預定使用量進行記錄和計數，避免了為 反復重新配置縮放參數而引起的中斷和延遲，大大提离了 圖像縮放處理速度。 本發明的上述任何實施例中，對於圖像縮放操作的演 算法並不做限定。當採用不同的演算法處理圖像縮放時， 只需要相應地改變圖像縮放所需要的縮放參數的類型，而 對於硬體本身並不需要進行變更，諸如縮放參數暫存器和 200945879 計數暫存器等可以預先設置’再在實際應用中採用相應數 量的暫存器’也可以直接複用電路中現有的暫存器。因此， 本發明可以作為通用的圖像縮放處理的硬體架構。例如， 即使在多維的圖像縮放處理中，只需改變縮放參數暫存器 中縮放參數的定義，而不必改變硬體結構，即可進行多維 的圖像縮放處理。本發明並不限於實施例所做的闡述，任 何基於本發明的修改和本發明的等同物都應涵蓋在本發明 的權利要求的精神和範圍之内 雖然本發明已以一較佳實施例揭露如上，然其並非用 以限定本發明，任何熟習此技藝者，在不脫離本發明之精 神和範圍内，當可作各種之更動與潤飾，因此本發明之保 護範圍當視後附之申請專利範圍所界定者為準。 【圖式簡單說明】 為讓本發明之上述和其他目的、特徵、優點與實施例 能更明顯易懂，所附圖式之詳細說明如下： 第1圖為先前技術中一種圖像縮放裝置的非線性縮放 的縮放參數分配原理示意圖； 第2A圖係為本發明之一種通用圖像縮放裝置之一結 構示意圖； 第2B圖係為本發明之另一種通用圖像縮放裝置之一 結構示意圖； 第3圖係為根據本發明的一個實施例的濾波器之一結 構示意圖； 21 200945879 _ 圖係為本發明之一實施例的濾波器的插值原理之 一示意圖； 第5圖係為本發明之—實施例之通用圖像縮放裝置 ’計數單元與控制單元的縮放參數暫存器之一對應關係 示意圖； 第6圖係為本發明之通用圖像縮放方法之一流程圖； 第7 ^圖係為本發明之一實施例之通用圖像縮放裝置 彳數早70與控制單元的縮放參數暫存器之—對應關係 髎 示意圖； 。第8圖係為本發明之一實施例之設置兩組縮放參數暫 存器之控制單元之示意圖；以及 第9圖為根據本發明該實施例的縮放參數和縮放參數 的預定使用量的分佈原理示意圖。 【主要元件符號說明】 21 .圖像縮放單元 22 :控制單元 ❹ 縮放參數暫存器 225 — 1:第一縮放參數暫存器 225 — 2:第二縮放參數暫存器組 組 2252 :量化階數暫存器 2254 :濾波器階數暫存器 23 — 1 :第一計數器單元 23 1 :第一計數器 233 :第三計數器 235 :計數器 2 2 51 .渡波函數暫存器 2253 ··步進值暫存器 23 ·•計數單元 23 — 2:第二計數器單元 232 :第二計數器 234 :第四計數器 24 :預測配置單元 22 200945879 41 :目標圖像 411 :目標畫素點 42 :原圖像 421、422、423、424 :原畫素 43 :濾波器視窗 點 45 :濾波函數Figure 6 is a flow chart of the general image scaling method of the present invention. The sixth method comprises the following steps: S1 'configure at least - a group image scaling parameter and a predetermined usage amount corresponding to the group of scaling parameters; S2, performing image scaling processing on the pixel points on the image, and scaling The usage amount of the parameter has been completed is counted; Μ, it is judged whether the usage amount of the zoom parameter has been completed reaches the configured predetermined usage amount; if not, the program returns to step S2; if yes, proceeds to step §4, reconfiguring the next set of scaling parameters and The predetermined amount of usage corresponding to the set of scaling parameters. The above image scaling method may further comprise the step of predicting a scaling parameter of the next image based on a scaling parameter of the image after an image scaling operation is completed. The number of scaling parameter registers 225 and counters 235 can be set as desired, and several different configurations are described below in several embodiments. According to an embodiment of the invention, the counting unit comprises at least one counter in which a predetermined usage amount of a set of scaling parameters is recorded. As shown in the fifth 200945879 diagram, the scaling parameter register 225 sets a value of a set of scaling parameters, including, for example, a filtering function, a quantization order, a step value, a filter order, etc., and the set of scaling parameters is recorded in the counter 235. The value of the scheduled usage. For example, the set of scaling parameters needs to be scaled by using one pixel point of the target image. The initial value of the counter can be set to 1 〇〇. After each zooming process of a pixel point is completed, the value of the counter is decremented, so that until the scaling process of 100 pixel points is completed, the value of the counter becomes zero, and the counting unit will prompt the scaling parameter register. A new set of scaling parameters is loaded and the predetermined usage of the set of scaling parameters that are entered together is recorded in the counter. According to another embodiment of the present invention, a set of counters may also be provided in the counting unit 23. The number of counters corresponds to the number of scaling parameter registers in the scaling parameter register group 225, and a corresponding counter is respectively configured for each scaling parameter register, and the scaling parameter in the scaling parameter register can be stored. The scheduled usage. As shown in Fig. 7, the first counter 231 (the text in the box indicates the object counted by the counter, the same below) and the filter function register 2251 in the scaling parameter register group should be, m 232 corresponds to the quantization step register 2252 in the scaling parameter register group, the third counter 233 corresponds to the step value register 2253 in the scaling parameter register group, and the fourth counter 234 and the scaling parameter are temporarily stored. The filter order register 2254 in the group corresponds. According to another embodiment of the present invention, as shown in FIG. 8, two sets of scaling parameter registers are set in the control unit, and the parameter group register can be accessed in turn for each set of scaling parameter registers. And each set-counter, each time refresh-group zoom parameter buffer value of the parameter register, the predetermined usage amount of the set of zoom parameter values is simultaneously entered into the corresponding count 17 200945879. Specifically, the scaling parameter that is currently used for the scaling process may be stored in the first scaling parameter register group 225-1, and the predetermined usage amount of the first group of scaling parameters may be stored in the first counter unit 23-1. During the operation using the first set of scaling parameters, the scaling parameter required for the subsequent scaling operation is stored in the second scaling parameter register group 225-2 and the second set of scaling parameters are scheduled to be used in correspondence. The second counting unit 23-2. When the counting unit indicates that the predetermined usage amount of the current scaling parameter stored in the first group of scaling parameter registers has arrived, and the scaling ❿ parameter needs to be converted, the scaling parameter stored in the second scaling parameter register group is directly started. Thereafter, when the second set of scaling parameters are used for operation, the scaling parameter to be used is externally input to refresh the first scaling parameter register group, and the corresponding scaling parameter predetermined usage amount is input to the first counting unit. When the counting unit indicates that the predetermined usage amount of the scaling parameter stored in the second scaling parameter register group has arrived, when the scaling parameter needs to be transformed, the value of the first scaling parameter register group is directly re-acquired. In this way, the two sets of scaling parameter registers and the corresponding two sets of counters are rotated, and when a group of zoom parameters is operated, another set of scaling parameters can be loaded simultaneously, which speeds up the processing. According to another embodiment of the present invention, a plurality of sets of scaling parameter registers and a plurality of counters respectively corresponding thereto may be set. When scaling a certain line of the target image, all the scaling parameters and their corresponding pre-quantity usages can be recorded, and other rows of the same image can be scaled according to the settings of the row. Fig. 9 is a schematic diagram showing the distribution principle of the predetermined usage amount of the scaling parameter and the scaling parameter according to the embodiment of the present invention. Referring to Fig. 9, it is assumed that the first line in the image needs to be nonlinearly reduced. 18 200945879 The 'divided into 5 parts' each part has its corresponding scaling parameter and the predetermined usage amount of the scaling parameter. For example, the first set of scaling parameters in FIG. 9 and the predetermined usage amount of the first set of scaling parameters, the second set of scaling parameters and the predetermined usage amount of the second set of scaling parameters 'the third set of scaling parameters and the third set of scaling parameters The predetermined usage amount of the fourth group scaling parameter and the fifth group scaling parameter is predetermined usage amount 'the fourth group scaling parameter and the fourth group scaling parameter's predetermined usage amount. Since the predetermined usage amount of the scaling parameter is recorded in the counting unit, and the scaling parameter can be prompted to be replaced according to the predetermined usage amount, the boundary between the five parts can be flexibly set. 'Each part is no longer limited to being fixed. The size and the starting address can be configured to correspond to the corresponding scaling parameter register and counter for the five parts. For the following lines, if you still divide the same 5 parts and use the same scaling parameters respectively, such as lines 2 to 6 in Figure 9, you can still use the already configured scaling parameters and scaling parameters directly. The quantity ' does not have to reload the respective scaling parameters of the 5 parts, thus also increasing the processing speed. The division of the scaling parameters of each line can also be determined according to the continuous requirements. For example, the 7th Φ to 10th line in Fig. 9 can also be divided into 4 parts. Corresponding scaling parameter registers and counters can be configured for each of the four sections. For an image, the counting unit may also set a vertical direction counter in the vertical direction of the image for recording the predetermined usage amount of the plurality of sets of scaling parameters on a horizontal line in the vertical direction, and reaching the predetermined usage amount in the vertical direction. When prompted, update the zoom parameters. The predetermined amount of usage in the vertical direction can be in units of lines. For example, the multi-group scaling parameters of lines 1 to 6 in Figure 9 are the same, and the corresponding vertical direction usage is recorded in the vertical direction counter, and each line is zoomed in the vertical direction of the image (4), 200945879 Put the vertical direction. The value of the tens register is reduced by i. Until the value of the vertical direction register is reduced to zero, the counting unit indicates the predetermined use of the vertical scaling parameter, and the scaling parameter of the seventh line is restarted. In this way, in the case that the current row and the top row use the same scaling parameter, the action of reloading the predetermined usage amount of the scaling parameter and the scaling parameter may be omitted in the switching line and the line image scaling, thereby reducing the number of interruptions, and It saves the power consumption and time of loading the scaling parameters, and also achieves the purpose of flexibly dividing the image into a plurality of different parts and scaling the φ according to different scaling parameters. In one embodiment of the invention, the image scaling device can pre-configure all scaling parameters of an image. First, set multiple sets of scaling parameters for the horizontal line, and configure the predetermined usage amount corresponding to each set of scaling parameters in the corresponding counter; and then configure multiple vertical direction counters. Each vertical direction counting register is set according to the same set of scaling parameters. The number of horizontal lines applied is determined by the predetermined amount of use of the scaling parameter in the vertical direction. In the multi-window image scaling process, the zoom parameter register and the corresponding counter can be set individually for each window. The zoom parameter of the window can be pre-stored in the zoom parameter register, and the counting unit records and counts the predetermined usage amount of each window zoom parameter separately, thereby avoiding the interruption and delay caused by repeatedly reconfiguring the zoom parameter, greatly Off the image scaling processing speed. In any of the above embodiments of the present invention, the algorithm for the image scaling operation is not limited. When image scaling is handled by different algorithms, only the type of scaling parameters required for image scaling needs to be changed accordingly, and no changes are needed for the hardware itself, such as scaling parameter register and 200945879 count temporary storage. The device can be pre-set to 're-use the corresponding number of registers in the actual application' or directly multiplex the existing registers in the circuit. Therefore, the present invention can be used as a hardware architecture for general image scaling processing. For example, even in multi-dimensional image scaling processing, multi-dimensional image scaling processing can be performed by simply changing the definition of the scaling parameter in the scaling parameter register without changing the hardware structure. The present invention is not limited to the embodiments, and any modifications based on the present invention and equivalents of the present invention are intended to be included within the spirit and scope of the appended claims. As above, it is not intended to limit the invention, and any person skilled in the art can make various modifications and retouchings without departing from the spirit and scope of the invention. The scope is defined. BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features, advantages and embodiments of the present invention will become more <RTIgt; Schematic diagram of a scaling parameter allocation principle of nonlinear scaling; FIG. 2A is a schematic structural diagram of a general image scaling device of the present invention; FIG. 2B is a schematic structural diagram of another general image scaling device of the present invention; 3 is a schematic structural diagram of a filter according to an embodiment of the present invention; 21 200945879 _ is a schematic diagram of the interpolation principle of a filter according to an embodiment of the present invention; FIG. 5 is a view of the present invention - A schematic diagram of a correspondence between a counting unit and a scaling parameter register of the control unit of the general image scaling device of the embodiment; FIG. 6 is a flowchart of a general image scaling method of the present invention; A general image scaling apparatus according to an embodiment of the present invention has a mapping relationship between the number of images 70 and the scaling parameter register of the control unit. 8 is a schematic diagram of a control unit for setting two sets of scaling parameter registers according to an embodiment of the present invention; and FIG. 9 is a distribution principle of predetermined usage amounts of scaling parameters and scaling parameters according to the embodiment of the present invention. schematic diagram. [Main component symbol description] 21. Image scaling unit 22: Control unit 缩放 scaling parameter register 225 - 1: first scaling parameter register 225 - 2: second scaling parameter register group 2252: quantization step Number register 2254: filter order register 23-1: first counter unit 23 1 : first counter 233: third counter 235: counter 2 2 51. wave function register 2253 · step value The register 23 • the counting unit 23 - 2: the second counter unit 232 : the second counter 234 : the fourth counter 24 : the prediction configuration unit 22 200945879 41 : the target image 411 : the target pixel point 42 : the original image 421 , 422, 423, 424: original pixel 43: filter window point 45: filter function
23twenty three