TW574830B

TW574830B - A method and device transmitting and receiving video flow in network

Info

Publication number: TW574830B
Application number: TW91119947A
Authority: TW
Inventors: Jr-Lung Yang; Man-Fang Huang; Chen-Yuan Kuo
Original assignee: Jr-Lung Yang; Man-Fang Huang; Chen-Yuan Kuo
Priority date: 2002-09-02
Filing date: 2002-09-02
Publication date: 2004-02-01

Description

574830 五、發明說明（1) 發明領域本發明係有關於一種處理影像資料流的方法與裝置，特別是有關於一種於網路中收發影像流的方法及裝置。發明背景在分秒必爭，溝通頻繁的時代，影像會議系統使得公司人員可省去旅途奔波所造成的時間、金錢、員工生產力上的耗費，而能夠即時召開企業體内外區域性及國際性的會議，進行即時性面對面的雙向溝通及資料分享。然而近年來影像會議裝置的普及率並未見明顯成長，其主要原因在於其設備成本、傳輸頻寬、相容性等。目前市面上見到的影像會議系統的影像單元多僅能提供單點對單點的影像會議，其係採用Η · 3 2 0、Η. 3 2 3、 Η. 3 2 4等通訊協定，用以建立連線、控制連線、音訊及影像之編解碼等。若有三方以上需參與影像會議，則必須透過多點控制器（Multipoint Control Unit，MCU)來達成。一般MCU可以調撥顯示模式或連續顯示模式運作。調撥顯示模式僅能選取單一影像流，依據其音訊或主要調撥控制，將該影像流傳送至所有與會者。連續顯示模式MCU接收影像會議中所有與會者傳來的影像資料，將該等影像資料結合，產生單一混合影像，並將該混合影像傳回給各與574830 V. Description of the invention (1) Field of the invention The present invention relates to a method and device for processing image data stream, and more particularly to a method and device for transmitting and receiving image stream in a network. BACKGROUND OF THE INVENTION In an era where every minute counts and there is frequent communication, the video conference system allows company personnel to save time, money, and employee productivity costs caused by traveling, and can immediately hold regional and international conferences inside and outside the company. Real-time face-to-face two-way communication and data sharing. However, in recent years, the popularity of video conferencing devices has not seen a significant increase. The main reasons are the cost of equipment, transmission bandwidth, and compatibility. At present, the image units of video conference systems seen on the market can only provide single-point to single-point video conferences, which use communication protocols such as Η · 3 2 0, Η. 3 2 3, Η. 3 2 4 and so on. In order to establish a connection, control connection, audio and video codec, etc. If more than three parties need to participate in the video conference, it must be achieved through a Multipoint Control Unit (MCU). In general, the MCU can operate in display mode or continuous display mode. The dial display mode can only select a single video stream and send the video stream to all participants based on its audio or main dial control. The continuous display mode MCU receives the video data from all participants in the video conference, combines these video data to generate a single mixed image, and returns the mixed image to each

第5頁 574830Page 5 574830

五、發明說明（2) 會者。實際上MCU在進行影像混合時，常因為輪入影像編解碼格式相異、圖像格式相異、量化值不同，或去兮、— 〜$略设數個影像頻道之影像資料送達MCU的時間差異等因素，而使得該複數個輸入影像的混合有困難，或使得影像更新率心到限制。、又美國專利第5 4 5 3 7 8 0號揭示一種在編碼域整合數個$ 像流成單一影像流的方法，該方法只能適用於Η· 26 1之= 殊區塊群（G0B)結構，對於較新且較有效率的Η· 2 6 3則不缔上所迷，編碼域合成兩種。晝素域合成的方法係每一一〃旦^ \取夂成晝素，將這些晝素合成一晝面二珥〜像k解碼傳回給單點。此法將花費很高的面重新編碼再的編碼解碼將使畫質變差。編试二里及纪憶體，且兩次編碼的影像流去合成畫面，而不二5成的方法則是利用已在於可以降低運算量，也可保私其解碼成晝素。其優點於：、、父佳畫質。唯其困難點在 1 ·從每個單點送來的晝面有同步的晝面。、X ’使得難以整合這些非 2·從每個單點送來的晝面有不合這些不同型式的晝面。、旦面形式，使得難以整里化值，使得難以選擇轉換參數維持原有的精 3 ·從每個單點送來的晝面有不同的置化值而能使相關的離散式餘弦V. Description of invention (2) Participants. In fact, when the MCU is mixing the images, it is often because the encoding and decoding formats of the rotation images are different, the image formats are different, and the quantization values are different. Or, the time for the image data of several image channels to reach the MCU is slightly set. And other factors, which makes it difficult to mix the plurality of input images, or limits the image update rate. The U.S. Patent No. 5 4 5 3 7 8 0 discloses a method for integrating several $ image streams into a single image stream in the coding domain. This method can only be applied to Η · 26 1 = special block group (G0B) The structure, for the newer and more efficient Η · 2 6 3 is not intrigued, the coding domain is synthesized two kinds. The method of synthesizing the celestial domain is to take each one ^^ to form a celestial element, and synthesize these celestial elements into a diurnal surface and two dipoles ~ like k decoding and return to a single point. This method will cost a lot of time to re-encode and then encode and decode will degrade the image quality. We edited and tested the Erli and Jiyi bodies, and the two-encoded video streams were used to synthesize the pictures. Instead of using 50% of the method, we could reduce the amount of calculations and keep it decoded into daylight. Its advantages are as follows: The only difficulty is that 1 · The day and night surface sent from each single point has a synchronized day and night surface. , X ′ makes it difficult to integrate these non-2. The diurnal surface sent from each single point has a diurnal surface that does not fit these different types. The surface form makes it difficult to trim the value and makes it difficult to choose the conversion parameters to maintain the original precision. 3 The day surface sent from each single point has a different set value to enable the related discrete cosine.

574830 五、發明說明（3) 因此本發明人有鑑於此，經過不斷研究測試後始有本發明之產生。本發明提供一種在在網路中收發影像流的方法及裝置，可應用於現今常用的影像編碼標準，諸如 MPEG、H. 261、H. 2 6 3。本發明不受各個來源影像流的畫面到達的快慢、晝面形式、量化值的影響，整合後的影像流在解碼後將呈現出單一影像，此單一影像是各個來源影像的影像總合，且晝質不變，各個晝面更新的頻率與未整合前一致。574830 V. Description of the invention (3) Therefore, the inventor has taken this into consideration, and the invention came into being after continuous research and testing. The present invention provides a method and device for transmitting and receiving video streams in a network, which can be applied to video coding standards commonly used today, such as MPEG, H.261, H.263. The invention is not affected by the speed of arrival of each source image stream, the daytime form, and the quantization value. The integrated image stream will present a single image after decoding. This single image is the sum of the images of each source image, and The diurnal quality does not change, and the frequency of each diurnal update is the same as before the integration.

發明概要本發明之一目的係提供一種於網路中收發影像流的方法，其適用於影像會議系統收發影像資訊，其係應用於現今常用的影像編碼標準，使得能夠在編碼域合成畫面，以減少計算量並維持原有編碼的精度。SUMMARY OF THE INVENTION An object of the present invention is to provide a method for transmitting and receiving video streams in a network. The method is suitable for video conference systems to send and receive video information. Reduce the amount of calculation and maintain the accuracy of the original encoding.

本發明之另一目的係提供一種於網路中收發影像流的方法，其適用於影像會議系統收發影像資訊，其係不受各個來源影像流的晝面到達快慢、晝面形式或量化值的影響，且整合後的影像流在解碼後將呈現出單一影像，其中該單一影像是各個來源影像流的影像總合，且其晝質不變，各個畫面更新的頻率與未整合前一致。為達成本發明上述目的，本發明提出一種一種於網路Another object of the present invention is to provide a method for transmitting and receiving image streams on a network, which is suitable for image conference systems to send and receive image information. Impact, and the integrated image stream will show a single image after decoding, where the single image is the sum of the images of each source image stream, and its day quality is unchanged, and the frequency of each frame update is the same as before the integration. In order to achieve the above object of the invention, the present invention provides a network

第7頁 574830 五、發明說明（4) 中收發影像流的方法，其包括：於一網路介面接收一建立連線訊號，其中該建立連線訊號係來自第一影像源，其包含一對應於一編解碼格式的編解碼辨識訊號；將該編解碼辨識訊號儲存於一記憶體中；由該第一影像源接收部分影像封包流；將該部分影像封包流拆解為部分影像訊號；以該部分影像訊號製成一集成影像訊號，其中該集成影像訊號包含該編解碼格式；以該集成影像訊號合成一集成影像封包流，其係包含該編解碼格式；將該集成影像封包流傳送至該第一影像源。本發明亦提供實現上述方法之一種於網路中收發影像流的裝置，其係為一影像處理單元，其包括：一網路介面，其係用以從一影像源接收建立連線訊號及部分影像封包流並將一集成影像封包流傳送至該影像源，其中該建立連線訊號包含一編解碼格式；一記憶體，其包含一建立連線運算區，其係用以辨認及儲存該建立連線訊號及其編解碼格式；一封包驅動器，其係用以將該部分影像封包流拆解為部分影像訊號，並將一集成影像訊號合成為一集成影像封包流；一多點控制單元，其係用以將該部分影像訊號變為該集成影像訊號；其中，該封包驅動器係與該多點控制單元連結，該多點控制單元係與該記憶體連結，該網路介面係與該記憶體連結。為使熟悉該項技藝人士瞭解本發明之目的、特徵及功效，茲藉由下述具體實施例，並配合所附之圖式，對本發明詳加說明，說明如后：Page 7 574830 5. The method for transmitting and receiving image streams in the description of the invention (4) includes: receiving a connection establishment signal through a network interface, wherein the connection establishment signal is from a first image source and includes a corresponding A codec identification signal in a codec format; the codec identification signal is stored in a memory; a portion of the image packet stream is received by the first image source; the portion of the image packet stream is decomposed into a portion of the image signal; The partial image signal is made into an integrated image signal, wherein the integrated image signal includes the codec format; an integrated image packet stream is synthesized from the integrated image signal, which includes the codec format; and the integrated image packet stream is transmitted to The first image source. The present invention also provides a device for transmitting and receiving an image stream in a network, which is an image processing unit and includes: a network interface for receiving a connection establishment signal and a part from an image source. An image packet stream and transmitting an integrated image packet stream to the image source, wherein the connection establishment signal includes a codec format; a memory including a connection establishment operation area for identifying and storing the establishment Connection signal and its codec format; a packet driver, which is used to decompose part of the image packet stream into part of the image signal and synthesize an integrated image signal into an integrated image packet stream; a multi-point control unit, It is used to change part of the image signal into the integrated image signal. The packet driver is connected to the multi-point control unit, the multi-point control unit is connected to the memory, and the network interface is connected to the memory. Body link. In order to make those skilled in the art understand the purpose, features, and effects of the present invention, the following specific embodiments and the accompanying drawings are used to explain the present invention in detail, as described later:

第8頁 574830 五、發明說明（5) 發明詳細說明為充分揭露本發明，茲配合圖式詳細說明如下。圖二顯示本發明方法之流程圖。步驟2 0 1係於一網路介面接收一建立連線訊號，其中該建立連線訊號係來自第一影像源，其包含一對應於一編解碼格式的編解碼辨識訊號。步驟2 0 3係將該編解碼辨識訊號儲存於一記憶體中。步驟2 0 5係由該第一影像源接收部分影像封包流。步驟2 0 7 係將該部分影像封包流拆解為部分影像訊號。步驟2 0 9係以該部分影像訊號製成一集成影像訊號，其中該集成影像訊號包含該編解碼格式。步驟2 1 1係以該集成影像訊號合成一集成影像封包流，其係包含該編解碼格式。步驟2 1 3 係將該集成影像封包流傳送至該第一影像源。圖三顯示圖二所示本發明方法之示意圖。將每一個晝面置於一指定位置，再將其他三個位置放入而不編碼。如此雖然每次只更新四分之一的畫面，但當更新頻率夠高時，將呈現四個子晝面連續顯現的效果。每個子晝面依其晝面到達的頻率更新其晝面，快的影像流將不受制於慢的影像流，且其晝質與處理前一樣。第四圖顯示依據本發明之影像會議系統方塊圖。影像會議系統4 0 0包含·影像資料收發裝置4 5 0及兩個與會者端裝置430(在此包含第一裝置410及第二裝置42 0)。其中該與會者端裝置之數量係可以依據需要而增加。Page 8 574830 V. Description of the invention (5) Detailed description of the invention In order to fully disclose the present invention, the detailed description with the drawings is as follows. Figure 2 shows a flowchart of the method of the present invention. Step 201 is receiving a connection establishment signal through a network interface, wherein the connection establishment signal is from a first image source and includes a codec identification signal corresponding to a codec format. Step 203 stores the codec identification signal in a memory. Step 205 is receiving a partial image packet stream from the first image source. Step 2 0 7 is to disassemble the partial image packet stream into partial image signals. Step 209 is to make an integrated image signal from the partial image signal, where the integrated image signal includes the codec format. Step 2 1 1 uses the integrated video signal to synthesize an integrated video packet stream, which includes the codec format. Step 2 1 3 is transmitting the integrated image packet stream to the first image source. FIG. 3 shows a schematic diagram of the method of the present invention shown in FIG. 2. Place each day in a designated position, and place the other three positions without coding. Although only one quarter of the screen is updated at a time, when the update frequency is high enough, the effect of continuous display of four sub-day surfaces will appear. Each sub-day surface updates its day surface according to the frequency of its arrival. The fast image stream will not be restricted by the slow image stream, and its day quality is the same as before processing. The fourth figure shows a block diagram of a video conference system according to the present invention. The video conference system 400 includes a video data transmitting and receiving device 450 and two participant-side devices 430 (including a first device 410 and a second device 420). The number of participant devices can be increased as needed.

574830 五、發明說明（6) 影像資料收發裝置4 5 0係可接收來自複數個影像輸入頻道的影像資料，其亦可連續傳送個別影像資料，其中該影像資料之傳送係於複數個輸出影像流進行。第一裝置4 1 0係用以接收該等輸出影像流中之一輸出影像流，並產生一影像影像，且該影像影像包含來自該等影像輸入頻道的圖像。第二裝置4 2 0係用以接收該等輸出影像流中之另一輸出影像流並產生另一影像影像，且該影像影像包含來自該等影像輸入頻道的圖像。574830 V. Description of the invention (6) The image data transmitting and receiving device 450 can receive image data from a plurality of image input channels, and it can also continuously transmit individual image data, wherein the transmission of the image data is in a plurality of output image streams get on. The first device 4 10 is configured to receive one of the output image streams and generate an image image, and the image image includes images from the image input channels. The second device 4 2 0 is used to receive another output image stream in the output image streams and generate another image image, and the image image includes images from the image input channels.

圖五顯示依據本發明之影像資料收發裝置5 0 0，其包括：一網路介面5 0 1、一記憶體5 0 3、一封包驅動器5 0 5、一多點控制單元5 0 7。其中，封包驅動器5 0 5係與多點控制單元5 0 7連結，多點控制單元5 0 7係與記憶體5 0 3連結，網路介面5 0 1係與記憶體5 0 3連結。網路介面5 0 1係用以從一影像源接收建立連線訊號及部分影像封包流，並將一集成影像封包流傳送至該影像源，其中該建立連線訊號包含一編解碼格式。FIG. 5 shows a video data transmitting and receiving device 500 according to the present invention, which includes: a network interface 501, a memory 503, a packet driver 505, and a multipoint control unit 507. Among them, the packet driver 505 is connected to the multipoint control unit 507, the multipoint control unit 507 is connected to the memory 503, and the network interface 501 is connected to the memory 503. The network interface 501 is used for receiving a connection establishment signal and a partial image packet stream from an image source, and transmitting an integrated image packet stream to the image source, wherein the connection establishment signal includes a codec format.

記憶體5 0 3係包含一建立連線運算區5 0 4，其係用以辨認及儲存該建立連線訊號及其編解碼格式。封包驅動器5 0 5係用以將該部分影像封包流拆解為部分影像訊號，並將一集成影像訊號合成為一集成影像封包流。多點控制單元5 0 7係用以將該部分影像訊號變為該集The memory 503 includes a connection establishment operation area 504, which is used to identify and store the connection establishment signal and its codec format. The packet driver 50 is used to decompose the partial image packet stream into a partial image signal and synthesize an integrated image signal into an integrated image packet stream. Multi-point control unit 5 0 7 is used to change part of the video signal into the set

第10頁 574830 五、發明說明（7) 成影像訊號。第六圖顯示本發明實施例之一種於網路中收發影像流的方法之流程圖。步驟6 0 1係為於該網路介面接收第二建立連線訊號，其中該第二建立連線訊號係來自第二影像源，其包含一對應於第二編解碼格式的第二編解碼辨識訊號。步驟6 0 3係為將該第二編解碼辨識訊號儲存於該記憶體中。步驟6 0 5係為由該第二影像源接收第二部分影像封包流。步驟6 0 7係為將該第二部分影像封包流拆解為第二部分影像訊號。步驟6 0 9係為以該第二部分影像訊號製成一第二集成影像訊號，其中該第二集成影像訊號包含該第二編解碼格式。步驟6 1 1係為以該第二集成影像訊號合成第二集成影像封包流，其係包含該第二編解碼格式。步驟 6 1 3係為以該第二集成影像訊號合成第三集成影像封包流，其係包含該編解碼格式。步驟6 1 5係為將該第二集成影像封包流傳送至該第二影像源。步驟6 1 7係為將該第三集成影像封包傳送至該第一影像源。茲以實施例說明如下。圖一顯示一視訊會議系統中之多點控制單元所達成功能示意圖。設若一影像會議中有四名與會者，且該四名與會者各自藉由其影像會議設備產生影像資料，並將該等影像資料傳送至一多點控制單元（MCU)。本發明方法一般係應用於三方以上之影像會議，並可依據實際與會人數加以調整應用。Page 10 574830 V. Description of the invention (7) Video signal. The sixth figure shows a flowchart of a method for transmitting and receiving video streams in a network according to an embodiment of the present invention. Step 601 is to receive a second connection establishment signal through the network interface, where the second connection establishment signal is from a second image source and includes a second codec identification corresponding to the second codec format. Signal. Step 603 is to store the second codec identification signal in the memory. Step 605 is receiving the second part of the image packet stream by the second image source. Step 607 is to disassemble the second image packet stream into the second image signal. Step 609 is to form a second integrated image signal from the second partial image signal, wherein the second integrated image signal includes the second encoding and decoding format. Step 6 1 1 is to synthesize a second integrated image packet stream with the second integrated image signal, which includes the second encoding and decoding format. Step 6 13 is to synthesize a third integrated image packet stream with the second integrated image signal, which includes the codec format. Step 6 15 is for transmitting the second integrated image packet stream to the second image source. Step 6 17 is for transmitting the third integrated image packet to the first image source. Examples are described below. Figure 1 shows the success of a multipoint control unit in a video conference system. Suppose there are four participants in a video conference, and the four participants each use their video conference equipment to generate video data and transmit the video data to a multipoint control unit (MCU). The method of the present invention is generally applied to video conferences with more than three parties, and can be adjusted and applied according to the actual number of participants.

574830 五、發明說明（8) 依據本發明方法之影像會議系統中，多點控制單元係可以產生輸出集成 CIF(common intermediate format)訊號。多點控制單元將該輸出集成C I F訊號分割為四個正交象限，亦即，右上（第一象限）、左上（第二象限）、左下（第三象限）、右下（第四象限）四個象限。每一與會者端之輸入頻道包含一部分QCIF(quarter-coinin〇n intermediate format)訊號。而每一頻道對應於該輸出集成C I F訊號之一象限。當一影像經由該等頻道之一傳入多點控制單元時，多點控制單元將該影像指定至該輸出集成 C I F訊號之某一特定象限。目前通用的編碼解碼格式有許多種，其均可以與本發明配合使用。以下將以目前影像會議最常用的影像編碼標準H· 2 61及H. 2 6 3為例，說明本發明方法之運作。參見圖τ 二’设右來自第一與會者之輸入訊號1 0係為Η 26 1格式，來自弟一與會者之輸入訊號11、來自第三與會者之輸入气號1 2、來自第四與會者之輸入訊號1 3係屬η · 2 6 3格式。因之，多點控制單元所產生之輸出訊號1 4係以Η· 2 6丨格式傳送給第一與會者，而以Η · 2 6 3格式傳送給第二與會者、第三與會者、第四與會者。依據本發明方法，多點控制單元每次只更新四分之一的輸出集成C I F訊號，因而可以簡化多點控制單元的結構及降低其運作所負荷的運算量。依據本發明方法，多點押制單元僅需處理編碼域資料，故其運算所需的記憶體亦ς 傳統方法為少。如上所述，依據本發明方法的多點控制單574830 V. Description of the invention (8) In the video conference system according to the method of the present invention, the multi-point control unit can generate an output integrated CIF (common intermediate format) signal. The multi-point control unit divides the output integrated CIF signal into four orthogonal quadrants, that is, upper right (first quadrant), upper left (second quadrant), lower left (third quadrant), lower right (fourth quadrant). Quadrants. Each participant's input channel contains a portion of a QCIF (quarter-coininon intermediate format) signal. Each channel corresponds to one quadrant of the output integrated CIF signal. When an image is transmitted to the multipoint control unit via one of the channels, the multipoint control unit assigns the image to a specific quadrant of the output integrated CIF signal. There are many types of encoding and decoding formats commonly used at present, all of which can be used in conjunction with the present invention. In the following, the most commonly used image coding standards H. 2 61 and H. 2 6 3 of the current video conference will be taken as examples to illustrate the operation of the method of the present invention. See Figure τII '. Set the input signal 1 0 from the first participant to Η 26 1 format, the input signal from the first participant 11, the input gas number 1 from the third participant 2, and the fourth participant The input signal 1 3 is in η · 2 6 3 format. Therefore, the output signal 1 4 generated by the multi-point control unit is transmitted to the first participant in the format of Η · 2 6 丨 and transmitted to the second participant, the third participant, and the first participant in the format of Η · 2 6 3. Four participants. According to the method of the present invention, the multipoint control unit updates only a quarter of the output integrated CIF signal at a time, so that the structure of the multipoint control unit can be simplified and the amount of calculations required for its operation can be reduced. According to the method of the present invention, the multi-point holding unit only needs to process the data of the coding domain, so the memory required for its operation is also small compared with the traditional method. As mentioned above, the multi-point control order according to the method of the present invention

574830 五、發明說明（9) 元在複雜度、運算負荷、記憶體需求上均較傳統方法大為減少，故可以實施於目前市面上任何點對點影像會議裝置，而無須另外設置一中心伺服器。如前文所述，在本實施例中有四名與會者參加一影像會議。第一與會者之影像會議裝置於一影像頻道中產生部分QC I F輸入訊號1 0。第二與會者之影像會議裝置於一影像頻道中產生部分QC I F輸入訊號1卜第三與會者之影像會議裝置於一影像頻道中產生部分QC I F輸入訊號1 2。第四與會者之影像會議裝置於一影像頻道中產生部分QC I F輸入訊號 1 3。其中，每一與會者端用以產生部分QC I F輸入訊號之攝影機、電腦、螢幕及其他影像裝置係可採用習知之設備，無庸贅述。多點控制單元接收部分QC I F輸入訊號1 0、部分QC I F輸入訊號1 1、部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3，並將之結合為輸出集成CIF訊號14(參見圖三）。其中每一個部分QC I F輸入訊號包含一數位影像流。部分QC I F輸入訊號1 0中的數位影像係用以更新C I F輸出訊號1 4之第二象限影像。部分QC I F輸入訊號1 1中的數位影像係用以更新 C l F輸出訊號1 4之第一象限影像。部分QC I F輸入訊號1 2中的數位影像係用以更新C I F輸出訊號1 4之第三象限影像。部分QCIF輸入訊號13中的數位影像係用以更新CIF輸出訊號1 4之第四象限影像。多點控制單元所產生的C I F輸出訊號中，每一個C I F輸出訊號中僅有一個象限的影像被更新。每個新的C I F輸出訊號所更新的影像象限，係可以依574830 V. Description of the invention (9) Yuan has greatly reduced the complexity, computing load and memory requirements compared to traditional methods, so it can be implemented in any point-to-point video conference device on the market without the need to set up a separate central server. As mentioned above, in this embodiment, four participants participate in a video conference. The video conference device of the first participant generates part of the QC I F input signal 10 in a video channel. The video conference device of the second participant generates part of the QC I F input signal 1 in an image channel. The video conference device of the third participant generates part of the QC I F input signal 12 in an image channel. The video conference device of the fourth participant generates a part of the QC I F input signal in a video channel. Among them, the cameras, computers, screens, and other imaging devices used by each participant to generate part of the QC I F input signal can use conventional equipment and need not be described in detail. The multi-point control unit receives part of the QC IF input signal 1 0, part of the QC IF input signal 1 1, part of the QC IF input signal 1 2, part of the QC IF input signal 1 3, and combines it into an output integrated CIF signal 14 (see figure three). Each part of the QC I F input signal contains a digital video stream. Part of the digital image in QC I F input signal 10 is used to update the second quadrant image of C I F output signal 14. Part of the digital image in the QC I F input signal 1 1 is used to update the first quadrant image of the C l F output signal 14. Part of the digital image in the QC I F input signal 12 is used to update the third quadrant image of the C I F output signal 1 4. Part of the digital image in the QCIF input signal 13 is used to update the fourth quadrant image of the CIF output signal 14. Of the C I F output signals generated by the multipoint control unit, only one quadrant image in each C I F output signal is updated. The updated image quadrant of each new C I F output signal can be

574830 、發明說明（ίο) 據需要而改，變，使得一個CIF輸出訊號中同時更新兩個以上象限的影像。上述輸入影像頻道係可以包含一次部分㈧丨F訊號 (sub Q C I F )，而該輸出訊號系可以包含一部分q c丨？訊號，其係包含該次部分QCIF訊號影像的一部份。直中，該次^分QCIF訊號影像的大小並非部分㈧”訊號影像之四分 ^八二f 一次部分QC 1 F訊號影像之部分解出後，仍可將該二二二，加入一部分QCIF影像。該輸入影像頻道係可以包 =訊號’❿該輸出訊號係可以為一 4Cif訊號。或號係像頻道係可以包含-4CIF訊號，而該輸出訊標準化格式jcr:”上述影像訊號係可以採用任何示模式，目，丨ί ί 若採用H.263格式之連續顯法亦可處母人傳达之影像訊號可以多達四個。本發明方入16個與會者的狀況，其係將4個圖像加 QCIFH # ，其中該4個CIF圖像係包含4個部分理mr4鬧像則包含該16名與會者的影像。同出影像之第—象卩P及繁-免大小的圖像。例如：4ci^ 而复第-：阳：ί限可以包☆2個輸入cif訊號， ”=象限及第四象限則包含8個部分qcif訊號。一夕點控制單元之部分QC 了 F於以比部分qCIf輸入訊號i i <1 ^新速度可574830, invention description (ίο) According to the need to change, change, so that a CIF output signal to update two or more quadrant images at the same time. The above-mentioned input image channel system may include a partial signal (sub Q C I F), and the output signal system may include a portion q c 丨? The signal is part of the image of the sub-QCIF signal. Straight, the size of the QCIF signal image is not a part of the ^ ". The quarter of the signal image is ^ 82 f. Once the part of the QC 1 F signal image is solved, you can still add the QCIF image to the 2222 The input image channel can include = signal '❿ The output signal can be a 4Cif signal. Or the image channel can include -4CIF signal, and the output signal standardized format jcr: "The above image signal system can use any Display mode, purpose, 丨 ί ί If the continuous display method in H.263 format can be used, the mother can send up to four video signals. The present invention addresses the situation of 16 participants by adding 4 images to QCIFH #, where the 4 CIF images contain 4 partial images. The mr4 alarm contains the images of the 16 participants. The first image of the same image-Elephant P and Fan-free size images. For example: 4ci ^ and the second-: Yang: ί quadrant can include ☆ 2 input cif signals, "= quadrant and the fourth quadrant contain 8 partial qcif signals. The QC part of the overnight control unit is FC to I Part of the qCIf input signal ii < 1 ^ new speed available

QdF輪入訊號13之更新速度匕入Γ虎i2、部分則’故輪入訊號更新速度的内忮亚不會影響多點控制單元The update speed of the QdF round-in signal 13 enters the Γ tiger i2, and some parts. Therefore, Neaia will not affect the multi-point control unit.

574830 五、發明說明（π)574830 V. Description of the invention (π)

的運作。當部分QC I F輸入訊號1 0、部分QC I F輸入訊號Π、部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3中任何一者送達多點控制單元時，多點控制單元即處理之，產生並傳送一 C I F輸出訊號1 4至每一與會者的影像會議裝置。當下一個輸入訊號送達時，多點控制單元立即處理之，產生並傳送一個新的C I F輸出訊號1 4至每一與會者的影像會議裝置。由於與會者端有一方係採用H. 26 1格式，另外三方係採用Η. 2 6 3格式，故多點控制單元接收部分QC I F輸入訊號 1 0、部分QC I F輸入訊號U、部分QC I F輸入訊號1 2、部分 QCIF輸入訊號1 3後所產生的CIF輸出訊號14，亦有Η. 261及 H. 2 6 3兩種格式。若部分QC I F輸入訊號1 0係採Η. 2 6 1格式，而C I F輸出訊號1 4採Η. 2 6 3格式，則多點控制單元接收Η. 2 6 1格式的部分 QC I F輸入訊號1 0後，必須將之轉換為Η. 2 6 3格式，以產生 Η. 2 6 3格式的CIF輸出訊號14。同理，若部分QCIF輸入訊號 11、部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3係採 Η · 2 6 3格式，而C I F輸出訊號1 4為Η. 2 6 1格式，則多點控制單元接收Η. 2 6 3格式的部分QC I F輸入訊號1卜部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3後，必須將之轉換為、Operation. When any of the QC IF input signal 10, some QC IF input signal Π, some QC IF input signal 1 2, some QC IF input signal 1 3 arrive at the multipoint control unit, the multipoint control unit processes it. Generate and send a CIF output signal 14 to the video conference device of each participant. When the next input signal arrives, the multipoint control unit processes it immediately, generating and transmitting a new CIFF output signal 14 to the video conference device of each participant. Since one of the participants uses the H. 26 1 format and the other three parties use the Η. 2 6 3 format, the multipoint control unit receives part of the QC IF input signal 10, part of the QC IF input signal U, and part of the QC IF input. Signal 1 2. The CIF output signal 14 generated after some QCIF input signals 1 3 are also available in Η. 261 and H. 2 6 3 formats. If part of the QC IF input signal 10 is in the 6. 2 6 1 format, and the CIF output signal is 1 in the 4 ’s. 2 6 3 format, the multi-point control unit receives the Η. 2 6 1 part of the QC IF input signal 1 After 0, it must be converted into Η. 2 6 3 format to generate CIF output signal 14 in Η. 2 6 3 format. Similarly, if part of the QCIF input signal 11, part of the QC IF input signal 1 2, part of the QC IF input signal 1 3 is in the Η · 2 6 3 format, and the CIF output signal 1 4 is in the Η. 2 6 1 format, more The point control unit receives Η. 2 6 3 part of the QC IF input signal 1 and part of the QC IF input signal 1 2. After the part of the QC IF input signal 1 3, it must be converted into,

Η. 2 6 1格式，以產生Η · 2 6 1格式的C I F輸出訊號1 4。首先以C I F輸出訊號為Η. 2 6 3格式為例。Η. 2 6 1 format to generate Η · 2 6 1 format C I F output signal 1 4. First take the C I F output signal as Η. 2 6 3 format as an example.

在此實施例中，設若多點控制單元接收並處理部分 QC I F輸入訊號1 0、部分QC I F輸入訊號1 1、部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3，據以產生一 Η. 2 6 3格式的C I FIn this embodiment, it is assumed that the multi-point control unit receives and processes part of the QC IF input signal 10, part of the QC IF input signal 1 1, part of the QC IF input signal 1 2, and part of the QC IF input signal 1 3, thereby generating a Η. 2 6 3 format CIF

第15頁 574830 五、發明說明（12) 輸出訊號1 4。並將C I F輸出訊號1 4傳送給第一與會者、第二與會者、第三與會者、第四與會者。多點控制單元負責接收部分Q C I F輸入訊號1 0、部分 QC I F輸入訊號1 1、部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3，等待該等輸入訊號傳入新的影像資料。部分QC I F輸入訊號1 0首先送達，傳入一新部分QC I F影像至多點控制單元。多點控制單元將該部分QC I F影像的格式由Η. 2 6 1轉換為Η. 2 6 3，其係藉由更改其表頭資料及編碼域資料達成。經更改之表頭資料將該影像資料標示為一 I NTER圖像（即Ρ 圖像）。多點控制單元保留該影像之編碼域資料，其係用以定義第一與會者端傳來的影像影像。多點控制單元可能會將該編碼域資料加以重組，但其所定義之影像並不會被多點控制單元所更動，或者僅稍做更動。多點控制單元繼而製作C I F輸出訊號1 4 Α (參見圖八 )。首先產生一 C I F圖像之表頭資料，其係描述該C I F影像種類及其編碼種類（I NTER)。繼之指定一時間參考值給該圖像。影像流中一系列圖像的時間參考值依序遞增1。如前文所述，部分QC I F輸入訊號1 0所傳入多點控制單元之影像資料係為C I F輸出訊號1 4之第二象限圖像，故多點控制單元將部分QC I F輸入訊號1 0所傳入之影像資料的編碼格式由Η. 2 6 1轉換為Η. 2 6 3，並將之置於C I F輸出訊號1 4之第二象限圖像。新傳入的部分QC I F輸入訊號係置於C I F輸出訊號之第二象限，由於CIF輸出訊號的移動向量資料（MVD)可能不Page 15 574830 V. Description of the invention (12) Output signal 14. The CIF output signal 14 is transmitted to the first participant, the second participant, the third participant, and the fourth participant. The multi-point control unit is responsible for receiving part of the Q C I F input signal 10, part of the QC I F input signal 1 1, part of the QC I F input signal 1 2, part of the QC I F input signal 1 3, and waiting for these input signals to transmit new image data. Part of the QC I F input signal 10 is first delivered, and a new part of the QC I F image is passed to the multi-point control unit. The multi-point control unit converts the format of this part of the QC IF image from Η. 2 6 1 to Η. 2 6 3, which is achieved by changing its header data and coding field data. The changed header data marks the image data as an I NTER image (ie, a P image). The multipoint control unit retains the coding domain data of the image, which is used to define the image image transmitted from the first participant. The multi-point control unit may reorganize the data in the coding domain, but the image defined by it will not be changed by the multi-point control unit, or only slightly changed. The multi-point control unit then produces a C I F output signal 1 4 Α (see Figure 8). First, a header data of a CIF image is generated, which describes the CIF image type and its encoding type (I NTER). This image is then assigned a time reference value. The time reference value of a series of images in the image stream is sequentially increased by one. As mentioned above, the image data of the QC IF input signal 10 transmitted to the multi-point control unit is the second quadrant image of the CIF output signal 14, so the multi-point control unit will input a part of the QC IF input signal 10 The encoding format of the incoming image data is converted from Η. 2 6 1 to Η. 2 6 3 and placed in the second quadrant image of the CIF output signal 1 4. The newly incoming part of the QC I F input signal is placed in the second quadrant of the C I F output signal. The motion vector data (MVD) of the CIF output signal may not be

第16頁 574830 五、發明說明（13) 同，故每一個區塊群（G0B)均需經過MVD調整。調整過區塊群之移動向量資料後，即與第一象限的各區塊集之資料連結（NOT C0D)，產生一個新的CIF G0B。同理，第三象限及第四象限的區塊集資料亦標示該區塊集内容未變更 (NOT C0D)。 C I F輸出訊號1 4 A經過上述處理後，呈現如圖八所示的狀況。當該訊號傳送給第二與會者、第三與會者、第四與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4 A之第二象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。多點控制單元將C I F輸出訊號1 4 A傳送給第二與會者、第三與會者、第四與會者之後，繼續監看輸入影像頻道是否有新的訊號傳入。判斷部分QC I F輸入訊號1 1是否傳入新資料。若部分QC I F輸入訊號1 1不含新資料，則多點控制單元繼續判斷其他輸入影像頻道是否包含新資料。若部分 QC I F輸入訊號1 1包含新資料，則進行下列程序。由於部分QC I F輸入訊號1 1之影像資料係為Η. 2 6 3格式，故無須進行格式轉換的動作。參見圖七，此影像係為I圖像，故多點控制單元需將其轉換為Ρ圖像。在區塊集層級加入C0D(coded macro block indi cat ion)資料，且區塊集種類係設定為INTR A，若其量化值（quant izer)經過調整，則將區塊集種類設定為INTRA + Q。區塊集的彩度設定值（MCBPC)由I圖像表格傳送至P圖像表格。亮度設定值（CBPY)則取其原始值之補Page 16 574830 V. Description of the invention (13) Same, so each block group (G0B) needs to be adjusted by MVD. After adjusting the motion vector data of the block group, it is linked with the data of each block set in the first quadrant (NOT C0D) to generate a new CIF G0B. Similarly, the block set data in the third and fourth quadrants also indicate that the content of the block set has not changed (NOT C0D). After the C I F output signal 1 4 A undergoes the above processing, it appears as shown in Figure 8. When the signal is transmitted to the second participant, the third participant, and the fourth participant, the video conference device on the participant side displays the image of the second quadrant of the CIF output signal 1 4 A on the video conference device on the participant side On the screen, and the images displayed in the other quadrants on the screen remain unchanged. After the multipoint control unit transmits the C I F output signal 1 4 A to the second participant, the third participant, and the fourth participant, it continues to monitor whether a new signal is input from the input video channel. It is judged whether part of the QC I F input signal 1 1 is transmitted with new data. If some QC I F input signals 1 1 do not contain new data, the multi-point control unit continues to determine whether other input video channels contain new data. If part of the QC I F input signal 1 1 contains new data, the following procedure is performed. Since some of the image data of QC I F input signal 1 1 is in the Η. 2 6 3 format, no format conversion action is required. Referring to Figure 7, this image is an I image, so the multipoint control unit needs to convert it to a P image. Add COD (coded macro block indi cation) data at the block set level, and set the block set type to INTR A. If the quantizer is adjusted, set the block set type to INTRA + Q . The chroma set value (MCBPC) of the block set is transferred from the I-picture table to the P-picture table. The brightness setting value (CBPY) is supplemented by its original value

第17頁 574830Page 17 574830

數（complement)。上述程序係用料及編碼域資料，其操作細節係々f而要變更其表頭資繼之，多點控制單元製作°技^° 示Η先產生一 CIF圖像之表頭：Π 14B(如圖八所像種類及其編碼種類（INTER)。繼二扣二係描述該CIF影本發明…影像流中一系列圖像數目。依據 1〇予夕j圖像的時間參考值依序遞增部分人訊號n所傳人多點Number (complement). The above program is used for materials and coding domain data. Its operation details are 々f and its header information needs to be changed. Then, the multi-point control unit produces the first display of the CIF image: Π 14B ( Figure 8 shows the type of image and its coding type (INTER). The second and second series describe the CIF film ... The number of series of images in the image stream. The time reference value of the image 10 is sequentially increased in part. There are more people transmitted by human signal n

係為C I F輸出訊號1 4之第一象限圄徐平兀之〜诼貝才叶却八^冢限圖像，故多點控制單元將 =分QCIF輸入訊號n所傳入之影像資料的由H 2 6 3格式的] 回像轉換為H· 2 6 3格式的P圖像，並將之置於CIF輸出訊號 1 4B之第一象限圖像。新傳入的部分Qc丨隱入訊號係置於 C I F輸出訊號之第一象限，由於c丨F輸出訊號的移動向量資料（MVD)可能不同，故每一個區塊群（G〇B)均需經過 M VD調整。調整過區塊群之移動向量資料後，即與第二象限的各區塊集之資料連結（其標示為Not C0D)，產生一個新的C I F GOB。同理，第三象限及第四象限的區塊集資料亦標示該區塊集内容未變更（NOT C0D)。It is the first quadrant of the CIF output signal 1 4 圄 Xu Pingwu ~ 诼 Beicai Ye Ye ^ tsuka limit image, so the multi-point control unit will = = QCIF input signal n from the image data passed in by H [2 6 3 format] echo is converted to P image in H · 263 format and placed in the first quadrant image of CIF output signal 1 4B. The newly incoming part Qc 丨 hidden signal is placed in the first quadrant of the CIF output signal. Since the motion vector data (MVD) of the c 丨 F output signal may be different, each block group (G0B) needs to be After M VD adjustment. After adjusting the motion vector data of the block group, it is linked with the data of each block set in the second quadrant (which is marked as Not C0D) to generate a new CIFF GOB. Similarly, the block set data in the third and fourth quadrants also indicate that the content of the block set is unchanged (NOT C0D).

C I F輸出訊號1 4B經過上述處理後，呈現如圖八所示的狀況。當該訊號傳送給第二與會者、第三與會者、第四與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4B之第一象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。After the C I F output signal 1 4B is processed as described above, the situation shown in Figure 8 is displayed. When the signal is transmitted to the second participant, the third participant, and the fourth participant, the video conference device on the participant side displays the image of the first quadrant of the CIF output signal 1 4B on the video conference device on the participant side. The image displayed on the screen and in the other quadrants on the screen remains unchanged.

第18頁 574830 五、發明說明（15) 多點控制單元將C I F輸出訊號1 4 B傳送給第二與會者、第三與會者、第四與會者之後，繼續監看輸入影像頻道是否有新的訊號傳入。判斷部分QC I F輸入訊號1 2是否傳入新資料。若部分Q C I F輸入訊號1 2包含新資料，則進行下列程序。由於部分QC I F輸入訊號1 2之影像資料係為IL 2 6 3格式，故無須進行格式轉換的動作。此影像係為P圖像，故多點控制單元毋需將其轉換為P圖像。繼之，多點控制單元製作C I F輸出訊號1 4 C (如圖八所示）。首先產生一 C I F圖像之表頭資料，其係描述該C I F影像種類及其編碼種類（INTER)。繼之指定一時間參考值給該圖像。該時間參考值係記錄未傳送的圖像數目。依據本發明方法，影像流中一系列圖像的時間參考值依序遞增 1。 C I F輸出訊號之第一象限及第二象限圖像之區塊集資料均含有值為1之NOT C0D標記。部分QCIF輸入訊號12所傳入多點控制單元之影像資料係為C I F輸出訊號1 4之第三象限圖像，故多點控制單元將部分QC I F輸入訊號1 2所傳入之影像資料的由Η. 2 6 3格式的I圖像轉換為Η. 2 6 3格式的P圖像，並將之置於C I F輸出訊號1 4 C之第三象限圖像。新傳入的部分QC I F輸入訊號係置於C I F輸出訊號之第三象限，由於C I F輸出訊號的移動向量資料（MVD)可能不同，故每一個區塊群（G0B)均需經過MVD調整。調整過區塊群之移動P.18 574830 V. Description of the invention (15) After the multi-point control unit transmits the CIF output signal 1 4 B to the second participant, the third participant, and the fourth participant, it continues to monitor whether there is a new input channel. Incoming signal. It is judged whether some QC I F input signals 1 2 are transmitted with new data. If part of the Q C I F input signal 1 2 contains new data, the following procedure is performed. Since some of the image data of QC I F input signal 12 is in IL 2 6 3 format, no format conversion action is required. This image is a P image, so the multi-point control unit does not need to convert it to a P image. Next, the multi-point control unit produces a C I F output signal 1 4 C (as shown in Figure 8). First, a header data of a C I F image is generated, which describes the type of the C I F image and its encoding type (INTER). A time reference is then assigned to the image. This time reference value records the number of untransmitted images. According to the method of the present invention, the time reference value of a series of images in the image stream is sequentially increased by one. The block set data of the first quadrant and second quadrant images of the C I F output signal both contain a NOT C0D flag with a value of 1. The image data of the multi-point control unit transmitted by part of the QCIF input signal 12 is the third quadrant image of the CIF output signal 1 4; therefore, the multi-point control unit will transfer the part of the image data transmitted by the QC IF input signal 12 Η. The I picture in 2 63 format is converted into the P picture in 2 2 63 format and placed in the third quadrant of the CIF output signal 1 4 C. Part of the newly input QC I F input signal is placed in the third quadrant of the C I F output signal. Since the motion vector data (MVD) of the C I F output signal may be different, each block group (G0B) needs to be adjusted by MVD. Adjusted block group movement

第19頁 574830 五、發明說明（16) 向量資料後，即與第四象限的各區 ^ 示為NOT C0D)，產生一個新的nF g〇、b。貝科連結（其標 CIF輸出訊號14C經過上述處理後，呈狀況。當該訊號傳送給第二與會者、盘备圖八所示的會者時，與會者端之影像會議^ θ者、第四與像顯示於該與會者端影像會議’裝置？=第而赏^其他象限所顯示的影像則未做變更。鸯幕上’ 第：：1 Ϊ制ί兀將CIF輸出訊號“C傳送給第二*合者、弟二興會者、第四盥合去與會者、否有新的訊號傳入？二斷邱八’繼續監看輸入影像頻道是資料。若QCIF輸入訊號13是否傳入新序。 Q_入讯娩13包含新資料，則進行下列程由於部分QC I F輸入訊號i 3之旦q多％妞尨头T 式，故無須谁 3之衫像賀枓係為Η· 2 6 3格 & 進仃格式轉換的動作。衫像係為p圖像Page 19 574830 V. Description of the invention (16) After the vector data, that is, the regions with the fourth quadrant (shown as NOT C0D), a new nF g0, b is generated. Beco Link (its standard CIF output signal 14C is in the condition after the above processing. When this signal is transmitted to the second participant, the participant shown in Figure 8 is prepared, the video conference at the participant side ^ θ, the first The fourth image is displayed on the participant's video conference 'installation? = No. and the other quadrants are not changed. On the screen' No. 1: 1 The system outputs the CIF output signal "C to The second * couple, younger couple, and fourth couple go to the meeting. Are there any new signals coming in? Second break Qiu Ba 'continue to monitor the input video channel is data. If QCIF input signal 13 is incoming The new sequence. Q_Into the birth 13 contains new information, then the following process is performed. Because some QC IF input signals are i 3 and q %% girl hoe style T, so no 3 shirts like He Yan are required. 2 6 3 grid & format conversion action. Shirt image is p image

圖像。巧像故夕點控制單元毋需將其轉換為P 繼之，夕示）。首先^點控制單元製作C 1 F輸出訊號1 4D (如圖八所像種類及复紙生一 CIF圖像之表頭資料，其係描述該CIF影給該圖像=種類（INTER)。繼之指定一時間參考值本發明方法，二間筝考值係記錄未傳送的圖像數目。依據 1〇 ’影像流中一系列圖像的時間參考值依序遞增 C I F輪'出却料均含有信。^之弟一象限及第二象限圖像之區塊集資 *、、、1之NOT C0D標記。部分QC丨F輸入訊號丨3所傳image. It happens that the point control unit does not need to be converted to P (see below). First, the control unit produces a C 1 F output signal 1 4D (as shown in Figure 8 and the header information of a CIF image, which describes the CIF image to the image = type (INTER). According to the method of the present invention, the two reference values of the present invention record the number of untransmitted images. According to the time reference value of a series of images in the image stream, the CIF wheel is sequentially incremented. Letter. ^ The brother of the first quadrant and the second quadrant image block fund raising * ,,, 1, NOT C0D mark. Part of the QC 丨 F input signal 丨 3

574830 五、發明說明（17) 入多點控制單元之影像資料係為C I F輸出訊號1 4之第四象限圖像，故多點控制單元將部分QC I F輸入訊號1 3所傳入之影像資料的由H. 2 6 3格式的I圖像轉換為H. 2 6 3格式的P圖像，並將之置於C I F輸出訊號1 4D之第四象限圖像。新傳入的部分QC I F輸入訊號係置於C I F輸出訊號之第四象限，由於CIF輸出訊號的移動向量資料（MVD)可能不同，故每一個區塊群（G0B)均需經過MVD調整。調整過區塊群之移動向量資料後，即與第三象限的各區塊集之資料連結（其標示為NOT C0D)，產生一個新的CIF G0B。574830 V. Description of the invention (17) The image data of the multi-point control unit is the fourth quadrant image of the CIF output signal 14, so the multi-point control unit will input some of the QC IF input signals 1 3 of the image data. The I image in the H. 2 63 format is converted to the P image in the H. 2 63 format and placed in the fourth quadrant image of the CIF output signal 1 4D. Part of the newly input QC I F input signal is placed in the fourth quadrant of the C I F output signal. Since the CIF output signal may have different motion vector data (MVD), each block group (G0B) needs to be adjusted by MVD. After adjusting the moving vector data of the block group, it is linked with the data of each block set in the third quadrant (labeled as NOT C0D) to generate a new CIF G0B.

C I F輸出訊號1 4D經過上述處理後，呈現如圖八所示的狀況。當該訊號傳送給第二與會者、第三與會者、第四與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4D之第四象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。多點控制單元將C I F輸出訊號1 4D傳送給第二與會者、第三與會者、第四與會者。再以輸出集成C I F訊號為Η. 2 6 1格式為例。After the C I F output signal 1 4D is processed as described above, the situation shown in Figure 8 is displayed. When the signal is transmitted to the second participant, the third participant, and the fourth participant, the video conference device on the participant side displays the image of the fourth quadrant of the CIF output signal 1 4D on the video conference device on the participant side. The image displayed on the screen and in the other quadrants on the screen remains unchanged. The multipoint control unit transmits the CIF output signals 1 4D to the second participant, the third participant, and the fourth participant. Take the output integrated C I F signal as Η. 2 6 1 format as an example.

在此實施例中，設若多點控制單元接收並處理部分 QC I F輸入訊號1 0、部分QC I F輸入訊號1卜部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3，據以產生一 Η. 2 6 1格式的C I F 輸出訊號1 4，並將C I F輸出訊號1 4傳送給第一與會者。多點控制單元負責接收部分Q C I F輸入訊號1 0、部分 QC I F輸入訊號1卜部分QC I F輸入訊號1 2、部分QC I F輸入訊號1 3，等待該等輸入訊號傳入新的影像資料。部分QC I F輸In this embodiment, it is assumed that the multi-point control unit receives and processes part of the QC IF input signal 10, part of the QC IF input signal 1 and part of the QC IF input signal 1 2, and part of the QC IF input signal 1 3, thereby generating a frame. . 2 6 1 format CIF output signal 14 and send CIF output signal 14 to the first participant. The multi-point control unit is responsible for receiving part of the Q C I F input signal 1 0, part of the Q C I F input signal 1 and part of the QC I F input signal 1 2. Part of the QC I F input signal 1 3, waiting for these input signals to transmit new image data. Partial QC I F lose

第21頁 574830 五、發明說明（18) 入訊號1 0首先送達，傳入一新部分Q C I F影像1 0 A至多點控制單元。由於部分QC I F影像1 0 A的格式為Η · 2 6 1，故多點控制單元毋須將該部分QC I F影像的格式轉換。多點控制單元繼而製作C I F輸出訊號1 4Ε (參見圖八 )。首先產生一 C I F圖像之表頭資料。繼之指定一時間參考值給該圖像。如前文所述，部分QC I F輸入訊號1 0所傳入多點控制單元之影像資料係為C I F輸出訊號1 4Ε之第二象限圖像，放多點控制單元將之置於C I F輸出訊號1 4Ε之第二象限圖像。部分QC I F影像資料之群組值可以依據需要調整，其係可以如圖十所示之方式變更。Η · 2 6 1格式的部分QC I F影像之群組值為1、3、5，其係與Η. 2 6 1格式的C I F影像第二象限圖像之群組值相吻合，故無須改變之。多點控制單元將C I F輸出訊號1 4Ε之第一象限圖像之 G0B表頭資料中的群組值設為2、4或6。第一象限每一個 G0B只有表頭資料但不含任何區塊集資料。同理，多點控制單元將C I F輸出訊號1 4E之第三象限圖像之G0B表頭資料中的群組值設為7、9或1 1。第三象限每一個G0B只有表頭資料但不含任何區塊集資料。多點控制單元將C I F輸出訊號1 4E之第四象限圖像之G0B表頭資料中的群組值設為8、 1 0或1 2。每一個第四象限G0B只有表頭資料但不含任何區塊集資料。若設有適當群組值之G0B表頭資料不但不含任何區塊集資料時，即標示為省略MBA，其係表示該影像會議裝置在接收某一 C I F輸出訊號1 4E時並未更新該象限之圖Page 21 574830 V. Description of the invention (18) The incoming signal 10 is first delivered, and a new part Q C I F image 10 A to multi-point control unit is transmitted. Since the format of some QC I F images 10 A is Η · 2 6 1, the multi-point control unit does not need to convert the format of some QC I F images. The multi-point control unit then produces a C I F output signal 1 4E (see Figure 8). First, a header data of a CIF image is generated. A time reference is then assigned to the image. As mentioned above, the image data of the multi-point control unit passed in by some QC IF input signal 10 is the second quadrant image of CIF output signal 1 4E. The multi-point control unit puts it in the CIF output signal 1 4E. The second quadrant image. The group value of some QC I F image data can be adjusted as needed, and it can be changed as shown in Figure 10. Η · The group values of some QC IF images in the 2 6 1 format are 1, 3, and 5, which are consistent with the group values of the 2 quadrant images in the 6. 2 6 1 format CIF image, so there is no need to change them. . The multipoint control unit sets the group value in the G0B header data of the first quadrant image of the C I F output signal 1 4E to 2, 4, or 6. Each G0B in the first quadrant only has header data but does not contain any block set data. Similarly, the multipoint control unit sets the group value in the G0B header data of the third quadrant image of the C I F output signal 1 4E to 7, 9, or 11. Each G0B in the third quadrant only has header data but does not contain any block set data. The multipoint control unit sets the group value in the G0B header data of the fourth quadrant image of the C I F output signal 1 4E to 8, 10, or 12. Each fourth quadrant G0B only has header data but does not contain any block set data. If the G0B header data with appropriate group values does not contain any block set data, it is marked as omitting MBA, which means that the video conference device did not update the quadrant when receiving a CIF output signal 1 4E Picture

第22頁 574830 五、發明說明（19) 像。 C I F輸出訊號1 4E經過上述處理後，呈現如圖九所示的狀況。當該訊號傳送給第一與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4E之第二象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。多點控制單元將C I F輸出訊號1 4E傳送給第二與會者、第三與會者、第四與會者之後’繼績監看輸入影像頻道是否有新的訊號傳入。判斷部分QC I F輸入訊號1 1是否傳入新資料。若部分QC I F輸入訊號11包含新影像資料1 1 A，則進行下列程序。由於部分QC I F輸入訊號1 1之影像資料係為Η. 2 6 3格式，故須將其格式轉換為Η. 261。此影像可以為I圖像或Ρ 圖像，在上述格式轉換過程中並無差異。多點控制單元保留該影像之編碼域資料，其係用以定義第一與會者端傳來的影像影像。多點控制單元可能會將該編碼域資料加以重組，但其所定義之影像並不會被多點控制單元所更動，或者僅稍做更動。繼之，多點控制單元製作C I F輸出訊號1 4F (如圖九 )。首先產生一 C I F圖像之表頭資料，其係描述該C I F影像種類及其編碼種類（I NTER)。繼之指定一時間參考值給該圖像。部分QC I F輸入訊號1 1 Α傳入多點控制單元之影像資料係為C I F輸出訊號1 4F第一象限圖像，故多點控制單元將部Page 22 574830 V. Description of the invention (19) Image. After the C I F output signal 1 4E is processed as described above, the situation shown in Figure 9 is displayed. When the signal is transmitted to the first participant, the video conference device on the participant side displays the image of the second quadrant of the CIF output signal 1 4E on the screen of the video conference device on the participant side, and the other quadrants on the screen display The image is unchanged. The multi-point control unit transmits the C I F output signal 1 4E to the second participant, the third participant, and the fourth participant ', and then monitors the performance of the input video channel for a new signal. It is judged whether part of the QC I F input signal 1 1 is transmitted with new data. If part of the QC I F input signal 11 contains new image data 1 1 A, the following procedure is performed. As part of the QC I F input signal 1 1 image data is in Η. 2 6 3 format, it must be converted to Η. 261. This image can be an I image or a P image, and there is no difference in the format conversion process described above. The multi-point control unit retains the coding domain data of the image, which is used to define the image image transmitted from the first participant. The multi-point control unit may reorganize the coding domain data, but the image defined by it will not be changed or only slightly changed by the multi-point control unit. Next, the multi-point control unit produces a C I F output signal of 1 4F (see Figure 9). First, a header data of a CIF image is generated, which describes the CIF image type and its encoding type (I NTER). This image is then assigned a time reference value. Part of the QC I F input signal 1 1 Α image data transmitted to the multi-point control unit is C I F output signal 1 4F first quadrant image, so the multi-point control unit will

第23頁 574830 五、發明說明（20) 分QC I F輸入訊號1 1 A傳入之影像資料的由Η. 2 6 3格式轉換為 Η. 2 61格式，並將之置於CIF輸出訊號14F第一象限圖像。部分QC I F影像資料之群組值可以依據需要調整，其係可以如圖十所示之方式變更。由於該部分QC I F係置於C I F輸出影像之第一象限，故需將其群組值由1、3、5改變為2、 4、6。如圖十所示，C I F輸出影像之第一象限的群組值必須為2、4、6。多點控制單元將C I F輸出訊號1 4F之第二象限圖像之 G0B表頭資料中的群組值設為1、3或5。第一象限每一個 GOB只有表頭資料但不含任何區塊集資料。同理，多點控制單元將C I F輸出訊號1 4F之第三象限圖像之GOB表頭資料中的群組值設為7、9或1 1。第三象限每一個GOB只有表頭資料但不含任何區塊集資料。多點控制單元將C I F輸出訊號1 4F之第四象限圖像之G0B表頭資料中的群組值設為8、 1 0或1 2。每一個第四象限GOB只有表頭資料但不含任何區塊集資料。若設有適當群組值之GOB只有表頭資料但不含任何區塊集資料，即標示為省略MBA，其係表示該影像會議裝置在接收某一 C I F輸出訊號1 4F時並未更新該象限之圖像。 C I F輸出訊號1 4F經過上述處理後，呈現如圖九所示的狀況。當該訊號傳送給第一與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4F之第一象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。Page 23 574830 V. Description of the invention (20) The QC IF input signal 1 1 A is converted from 传入. 2 6 3 format to Η. 2 61 format and placed in the CIF output signal 14F section One quadrant image. The group value of some QC IF image data can be adjusted as needed, and it can be changed as shown in Figure 10. Since this part of QC I F is placed in the first quadrant of the C I F output image, its group value needs to be changed from 1, 3, and 5 to 2, 4, and 6. As shown in Figure 10, the group value of the first quadrant of the C I F output image must be 2, 4, and 6. The multipoint control unit sets the group value in the G0B header data of the second quadrant image of the C I F output signal 1 4F to 1, 3, or 5. In the first quadrant, each GOB only has header data but does not contain any block set data. Similarly, the multipoint control unit sets the group value in the GOB header data of the third quadrant image of the C I F output signal 1 4F to 7, 9, or 11. In the third quadrant, each GOB only has header data but does not contain any block set data. The multipoint control unit sets the group value in the G0B header data of the fourth quadrant image of the C I F output signal 1 4F to 8, 10, or 12. Each fourth quadrant GOB only has header data but does not contain any block set data. If the GOB with the appropriate group value has only the header data but does not contain any block set data, it is marked as omitting MBA, which means that the video conference device did not update the quadrant when receiving a CIF output signal 1 4F Of images. After the C I F output signal 1 4F undergoes the above processing, it will appear as shown in Figure 9. When the signal is transmitted to the first participant, the video conference device on the participant side displays the image of the first quadrant of the CIF output signal 1 4F on the screen of the video conference device on the participant side, and the other quadrants on the screen display The image is unchanged.

第24頁 574830 五、發明說明（21) 多點控制單元將CIF輸出訊號Uf傳送給第三與會者:第四與會者之後，繼續監看輪入^像，者、否有新的訊號傳入。判斷部分QC丨F輸入訊號了項逭是資料。若部分QCIF輸入訊號12包含新影像；=傳入新行下列程序。 t丄Z A，則進輸入訊號11之影像資料係為H. 2⑽夂式’故f將，、格式轉換為H. 26卜此影像 ^ 圖像，在上述格式轉換過程中並無差旦。為1圖像或P 留該影像之編碼域資料，其係用以定義第一盥^ ^旱元保的影”像二多點控制單元可能會將該編碼域士，傳來組，但，、所定義之影像並不會被多點控元所更^以重者僅稍做更動。所更動，或點控制單元製作CIF輪出訊號UG( )。百先產生一CI_像之表頭資料。繼圖九考值給該圖像。相疋符間參部分QCIF輪入訊號12A傳入多點控元係為hf輸出訊號14G第三象限圖像’故多像資料分QCIF輸人訊號12A傳入之影像資料的由& 2^格^將部 H. 261格式，並將之置於nF輸出訊工轉換為影像資料之群組值可以依據需要調-整象以^ 影像之第Λ 部分扣敗置於⑽輸出 9、1卜如圖十所示，_出影像14G之第三象組 574830 五、發明說明（22) 值必須為7、9、1 1。多點控制單元將C I F輸出訊號1 4 G之第二象限圖像之 G0B表頭資料中的群組值設為1、3或5。其中每一個G0B只有表頭資料但不含任何區塊集資料。同理，多點控制單元將C I F輸出訊號1 4G之第一象限圖像之G0B表頭資料中的群組值設為2、4或6。第一象限每一個G0B只有表頭資料但不含任何區塊集資料。多點控制單元將C I F輸出訊號1 4G之第四象限圖像之G0B表頭資料中的群組值設為8、1 0或1 2。每一個第四象限G0B只有表頭資料但不含任何區塊集資料。若設有適當群組值之G0B只有表頭資料但不含任何區塊集資料，即標示為省略MBA，其係表示該影像會議裝置在接收某一 C I F輸出訊號1 4G時並未更新該象限之圖像。 12 61格式的（：1?輸出訊號146經過上述處理後，呈現如圖九所示的狀況。當該訊號傳送給第一與會者時，與會者端之影像會議裝置將C I F輸出訊號1 4G之第三象限的影像顯示於該與會者端影像會議裝置之螢幕上，而螢幕上其他象限所顯示的影像則未做變更。多點控制單元將C I F輸出訊號1 4G傳送給第二與會者、第三與會者、第四與會者之後，繼續監看輸入影像頻道是否有新的訊號傳入。判斷部分QC I F輸入訊號1 3是否傳入新資料。若部分QC I F輸入訊號1 3包含新影像資料1 3 A，則進行下列程序。Page 24 574830 V. Description of the invention (21) The multi-point control unit transmits the CIF output signal Uf to the third participant: after the fourth participant, continue to monitor the rotation image, if any new signals are coming in . It is judged that some QC 丨 F input signals are items. It is data. If part of QCIF input signal 12 contains new image; t 丄 Z A, the image data of the input signal 11 is H. 2⑽ 夂 format, so f will convert the format to H. 26. This image ^ image, there is no difference in the format conversion process. The coding field data of the image is saved for 1 image or P, which is used to define the shadow of the first image. Like the two multi-point control unit, the coding field may be transmitted to the group, but, The defined image will not be changed by the multi-point control element. ^ The most important one will only be slightly changed. The changed or point control unit will make the CIF wheel output signal UG (). Baixian will generate a CI_image header. The data is given to the image following the evaluation in Figure 9. Corresponding intersymbol part QCIF wheel input signal 12A is passed to the multi-point control system for hf output signal 14G third quadrant image 'so multi-image data is divided into QCIF input signal The 12A incoming image data is formatted by & 2 ^^^ H.261, and placed in the nF output signal. The group value converted to the image data can be adjusted as needed-the whole image to ^ the first Λ is partially deducted and placed on the output 9 and 1. As shown in Figure 10, the third image group of the 14G image 574830 5. The description of the invention (22) must be 7, 9, 1 1. The multi-point control unit will The group value in the G0B header data of the second quadrant image of the CIF output signal 1 4 G is set to 1, 3, or 5. Each G0B has only the header data but not the Contains any block set data. Similarly, the multipoint control unit sets the group value in the G0B header data of the first quadrant image of the CIF output signal 1 4G to 2, 4, or 6. Each G0B in the first quadrant Only the header data but not any block set data. The multipoint control unit sets the group value in the G0B header data of the fourth quadrant image of the CIF output signal 1 4G to 8, 10, or 12. Each A fourth quadrant G0B has only header data but does not contain any block set data. If G0B with the appropriate group value has only header data but does not contain any block set data, it is marked as omitting MBA, which means that The video conference device did not update the image of the quadrant when receiving a CIF output signal 1 4G. 12 61 format (: 1? Output signal 146 after the above processing, the situation shown in Figure 9 appears. When the signal When transmitting to the first participant, the video conferencing device on the participant's side displays the image of the third quadrant of the CIF output signal 1 4G on the screen of the video conferencing device on the participant's side, and the images displayed on the other quadrants on the screen No change. Multipoint control unit will be C After the IF output signal 1 4G is transmitted to the second participant, the third participant, and the fourth participant, continue to monitor whether a new signal is input from the input video channel. Determine whether some QC IF input signals 1 3 are transmitted with new data. If part of the QC IF input signal 1 3 contains new image data 1 3 A, perform the following procedure.

由於部分QC I F輸入訊號1 1之影像資料係為H. 2 6 3格式，故須將其格式轉換為Η. 2 6 1。此影像可以為I圖像或PSince the image data of some QC I F input signals 1 1 is in H. 2 6 3 format, it must be converted to Η 2 6 1. This image can be I image or P

第26頁 574830 五、發明說明（23) 圖像，在上述格式轉換過程中並無差異。多點控制單元保留該影像之編碼域資料，其係用以定義第一與會者端傳來的影像影像。多點控制單元可能會將該編碼域資料加以重組，但其所定義之影像並不會被多點控制單元所更動i或者僅稍做更動。繼之，多點控制單元製作C I F輸出訊號1 4H (如圖九所示）。首先產生一 C I F圖像之表頭資料。繼之指定一時間參考值給該圖像。部分QC I F輸入訊號1 3 A傳入多點控制單元之影像資料係為C I F輸出訊號1 4 Η第四象限圖像，故多點控制單元將部分QC I F輸入訊號1 3 Α傳入之影像資料的由Η. 2 6 3格式轉換為 H. 2 61格式，並將之置於CIF輸出訊號14H第四象限圖像。部分QC I F影像資料之群組值可以依據需要調整，其係可以如圖十所示之方式變更。由於該部分QC I F係置於C I F輸出影像之第四象限，故需將其群組值由1、3、5改變為8、 1 0、1 2。如圖十所示，C I F輸出影像1 4G之第四象限的群組值必須為8、1 0、1 2。多點控制單元將C I F輸出訊號1 4H之第二象限圖像之 G0B表頭資料中的群組值設為1、3或5。其中每一個G0B只有表頭資料但不含任何區塊集資料。同理，多點控制單元將C I F輸出訊號1 4H之第一象限圖像之G0B表頭資料中的群組值設為2、4或6。第一象限每一個G0B只有表頭資料但不含任何區塊集資料。多點控制單元將C I F輸出訊號1 4H之第三象限圖像之G0B表頭資料中的群組值設為7、9或1 1。每Page 26 574830 5. Description of the invention (23) There is no difference in the image conversion process. The multi-point control unit retains the coding domain data of the image, which is used to define the image image transmitted from the first participant. The multi-point control unit may reorganize the coding domain data, but the image defined by it will not be changed or only slightly changed by the multi-point control unit. Subsequently, the multi-point control unit produces a C I F output signal 1 4H (as shown in Figure 9). First, a header data of a CIF image is generated. This is followed by assigning a time reference value to the image. Part of the QC IF input signal 1 3 A into the multi-point control unit is the CIF output signal 1 4 Η fourth quadrant image, so the multi-point control unit will send part of the QC IF input signal 1 3 Α into the image data It is converted from Η. 2 6 3 format to H. 2 61 format and placed in the CIF output signal 14H fourth quadrant image. The group value of some QC IF image data can be adjusted as needed, and it can be changed as shown in Figure 10. Since this part of QC I F is placed in the fourth quadrant of the C I F output image, its group value needs to be changed from 1, 3, and 5 to 8, 10, and 12. As shown in Fig. 10, the group value of the fourth quadrant of the C I F output image 1 4G must be 8, 10, 12. The multi-point control unit sets the group value in the G0B header data of the second quadrant image of the C I F output signal 1 4H to 1, 3, or 5. Each G0B only has header data but does not contain any block set data. Similarly, the multipoint control unit sets the group value in the G0B header data of the first quadrant image of the C I F output signal 1 4H to 2, 4, or 6. Each G0B in the first quadrant only has header data but does not contain any block set data. The multi-point control unit sets the group value in the G0B header data of the third quadrant image of the C I F output signal 1 4H to 7, 9, or 11. each

第27頁 574830 五、發明說明（24) 一個第三象限G0B之表頭資料並不和任何區塊集次同。若設有適當群組值之G 0 B只有表頭資料伯二二；、、了十*—不含任柯塊集資料，即標示為省略MBA，其係表示該影像合J ^ 在接收某一 CIF輸出訊號14H時並未更新該象限之θ圖像'置 Η· 261格式的CIF輸出訊號14Η經過上述處理後"，呈°玉如圖九所示的狀況。當該訊號傳送給第一盘合去士見曰茶叶，與奋者端之影像會議裝置將c I F輸出訊號1 4Η之第四象限的景^ ^ 顯示於該與會者端影像會議裝置之螢幕上，而榮幕上= 象限所顯示的影像則未做變更。熟悉此技藝者應該都瞭解，除了 · 2 6 1和Η 2 6 3之外，現存許多其他種類的編碼解碼格式，其均可以與本發明方^ 配合使用，用以在一具有複數個與會者的影像會議中，從複數個影像頻道以一次一個影像的方式接收影像資料。.攸上述實施例中，用以將部分qCIF資料從複數個與會者端傳送至多點控制單元之硬體，以及將C I F資料從多點控制單元傳送至該等與會者端裝置之硬體，均屬已知，”故二在此不再贅述。Page 27 574830 V. Description of the invention (24) The header data of a third quadrant G0B is not the same as any block set. If G 0 B with the appropriate group value is only the header data Bo 22; 、, the ten *-excluding any Ke block set data, it is marked as omitted MBA, which means that the image is being received J ^ When a CIF output signal 14H does not update the θ image of the quadrant, the setting of the 261 format CIF output signal 14Η is as shown in Fig. 9 after the above processing. When the signal was sent to the first set of tea to Shijianyu, the video conference device of the endurance side displayed the scene of the fourth quadrant of the c IF output signal 14 ^ ^ ^ on the screen of the video conference device of the participant , And the image displayed on the glory = quadrant is unchanged. Those who are familiar with this art should understand that besides · 2 6 1 and Η 2 6 3, there are many other types of encoding and decoding formats that can be used in conjunction with the present inventor ^ to have a number of participants In a video conference, video data is received from multiple video channels one image at a time. In the above embodiment, the hardware used to transmit part of the qCIF data from the plurality of participants to the multipoint control unit, and the hardware to transmit the CIF data from the multipoint control unit to the participants' devices. It's known, "so I won't repeat it here.

CIF訊號中除了傳送影像給影像會議與會者之外，亦可以傳送其他類型的資料，例如：影片剪輯、文件、報表、簡報資料等。圖十一顯示依據本發明方法的影像會議系統之例示。此系統包含一依據本發明裝置之影像資料收發裝置，其係裝設於一中央伺服器中。影像訊號藉由乙太網路在N個與會者端裝置之間流通。傳統的多點控制單元需要另外設置In addition to transmitting images to video conference participants, CIF signals can also transmit other types of data, such as movie clips, documents, reports, and presentation materials. FIG. 11 shows an example of a video conference system according to the method of the present invention. The system includes an image data transmitting and receiving device according to the present invention, which is installed in a central server. The image signal is circulated between N participant devices through the Ethernet. Traditional multi-point control units need additional settings

574830 五、發明說明（25) 一伺服器才能運作，本發明方法之多點控制單元之硬體需求甚低，故可以輕易地透過軟體更改而設於任何影像裝置上（如圖十二所示）。另外，本發明方法亦可以應用其他種類的傳輸媒介，例如：行動無線通訊網路、I SDN、纜線 /DSL、ATM網路、有線電話網路等。另外，如前文所述，此系統中之影像裝置係可以包含個人電腦、2. 5G/3G影像行動電話、筆記型電腦、個人數位助理、遊戲機等。574830 V. Description of the invention (25) Only a server can operate. The hardware requirements of the multi-point control unit of the method of the present invention are very low, so it can be easily installed on any imaging device through software modification (as shown in Figure 12). ). In addition, the method of the present invention can also be applied to other types of transmission media, such as: mobile wireless communication network, I SDN, cable / DSL, ATM network, wired telephone network, etc. In addition, as mentioned above, the imaging device in this system may include a personal computer, a 2.5G / 3G image mobile phone, a notebook computer, a personal digital assistant, a game console, and the like.

雖然本發明已以數個較佳實施例揭露如上，然其並非用以限定本發明，任何熟悉此技藝者，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，凡所做之各種更動與潤飾皆在本發明後附之申請專利範圍内。Although the present invention has been disclosed as above with several preferred embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make various modifications and decorations without departing from the spirit and scope of the present invention. Various modifications and retouches made are within the scope of the patent application attached to the present invention.

第29頁 574830 圖式簡單說明圖式簡單說明第一圖顯示一視訊會議系統中之多點控制單元所達成功能示意圖。第二圖顯示本發明提供之一種用於在影像會議系統收發影像資訊的方法之流程圖。第三圖顯示圖二所示本發明方法之示意圖。第四圖顯示依據本發明裝置系統方塊圖。第五圖顯示依據本發明裝置之影像資料收發裝置方塊圖。第六圖顯示本發明實施例之一種用於在影像會議系統收發影像資訊的方法之流程圖。第七圖顯示依據本發明方法以H· 2 6 3格式傳送圖像之資料内容。第八圖顯示依據本發明方法製作CI F輸出訊號。第九圖顯示依據本發明方法以H· 2 6 3格式傳送圖像之資料内容。第十圖顯示依據本發明方法部分QC I F影像資料群組值之變更。第十一圖顯示依據本發明方法的影像會議系統之例示。第十二圖顯示本發明裝置透過軟體更改而設於任何影像裝置上。主要元件編號Page 29 574830 Brief description of the diagram Brief description of the diagram The first diagram shows the function of the multipoint control unit in a video conference system. The second figure shows a flowchart of a method for transmitting and receiving image information in an image conference system provided by the present invention. The third diagram shows a schematic diagram of the method of the present invention shown in FIG. The fourth figure shows a block diagram of a device system according to the present invention. The fifth figure shows a block diagram of an image data transmitting and receiving device according to the device of the present invention. The sixth figure shows a flowchart of a method for transmitting and receiving image information in an image conference system according to an embodiment of the present invention. The seventh figure shows the data content of an image transmitted in H · 263 format according to the method of the present invention. The eighth figure shows the production of a CI F output signal according to the method of the present invention. The ninth figure shows the data content of an image transmitted in H · 263 format according to the method of the present invention. The tenth figure shows the change of the QC IF image data group value according to the method of the present invention. FIG. 11 shows an example of a video conference system according to the method of the present invention. The twelfth figure shows that the device of the present invention is installed on any image device through software modification. Major component numbers

574830 圖式簡單說明 100 QCIF圖像之群組值表 150 C IF圖像之群組值表 2 0 1接收建立連線訊號的步驟 2 0 3將編解碼辨識訊號儲存於記憶體中的步驟 2 0 5由第一影像源接收部分影像封包流的步驟 2 0 7將部分影像封包流拆解的步驟 2 0 9以部分影像訊號製成集成影像訊號的步驟 2 1 1以集成影像訊號合成集成影像封包流的步驟 2 1 3將集成影像封包流傳送至第一影像源的步驟 4 0 0影像會議系統 4 10第一裝置 4 2 0第二裝置 4 3 0與會者端裝置 4 5 0影像資料收發裝置 5 0 0影像資料收發裝置 5 01網路介面 5 0 3記憶體 5 0 4建立連線運算區 5 0 5封包驅動器 5 0 7多點控制單元 6 0 1接收第二建立連線訊號的步驟 6 0 3將第二編解碼辨識訊號儲存的步驟 6 0 5由第二影像源接收第二部分影像封包流的步驟 6 0 7將第二部分影像封包流拆解的步驟574830 Schematic description of 100 QCIF image group value table 150 C IF image group value table 2 0 1 Step 2 of receiving connection establishment signal 2 0 3 Step 2 of storing codec identification signal in memory 0 5 Step of receiving a partial image packet stream from a first image source 2 0 7 Step of disassembling a partial image packet stream 2 0 9 Step of making an integrated image signal from a partial image signal 2 1 1 Synthesizing an integrated image with an integrated image signal Step 2 1 3 of the packet flow Step of transmitting the integrated image packet stream to the first image source 4 0 0 video conference system 4 10 first device 4 2 0 second device 4 3 0 participant device 4 5 0 video data transmission Device 5 0 0 Image data transceiver device 5 01 Network interface 5 0 3 Memory 5 0 4 Establish connection calculation area 5 0 5 Packet driver 5 0 7 Multi-point control unit 6 0 1 Steps for receiving a second connection establishment signal 6 0 3 Step for storing the second codec identification signal 6 0 5 Step for receiving the second part of the image packet stream from the second image source 6 0 7 Step for disassembling the second part of the image packet stream

第31頁 574830 圖式簡單說明 6 0 9製成一第二集成影像訊號的步驟 6 1 1合成第二集成影像封包流的步驟 6 1 3合成第三集成影像封包流的步驟 6 1 5將第二集成影像封包流傳送至第二影像源的步驟 6 1 7將第三集成影像封包傳送至第一影像源的步驟Page 574830 Brief description of the diagram 6 0 9 Steps to make a second integrated image signal 6 1 1 Step to synthesize a second integrated image packet stream 6 1 3 Steps to synthesize a third integrated image packet stream 6 1 5 Steps of transmitting two integrated image packets to the second image source 6 1 7 Steps of transmitting third integrated image packets to the first image source

第32頁Page 32

Claims

574830 六、申請專利範圍 1. 一種於網路中收發影像流的方法，其包括： A. 於一網路介面接收一建立連線訊號，其中該建立連線訊號係來自第一影像源，其包含一對應於一編解碼格式的編解碼辨識訊號； B. 將該編解碼辨識訊號儲存於一記憶體中； C. 由該第一影像源接收部分影像封包流； D. 將該部分影像封包流拆解為部分影像訊號； E. 以該部分影像訊號製成一集成影像訊號，其中該集成影像訊號包含該編解碼格式； F. 以該集成影像訊號合成一集成影像封包流，其係包含該編解碼格式， G. 將該集成影像封包流傳送至該第一影像源。 2 .如申請專利範圍第1項所述之方法，其進一步包含：； A. 於該網路介面接收第二建立連線訊號，其中該第二建立連線訊號係來自第二影像源，其包含一對應於第二編解碼格式的第二編解碼辨識訊號； B. 將該第二編解碼辨識訊號儲存於該記憶體中； C. 由該第二影像源接收第二部分影像封包流； D. 將該第二部分影像封包流拆解為第二部分影像訊號； E. 以該第二部分影像訊號製成一第二集成影像訊號，其中該第二集成影像訊號包含該第二編解碼格式； F. 以該第二集成影像訊號合成第二集成影像封包流，其係包含該第二編解碼格式； G. 以該第二集成影像訊號合成第三集成影像封包流，其係574830 VI. Scope of patent application 1. A method for transmitting and receiving image streams in a network, including: A. receiving a connection establishment signal through a network interface, wherein the connection establishment signal is from a first image source, and Including a codec identification signal corresponding to a codec format; B. storing the codec identification signal in a memory; C. receiving a partial image packet stream from the first image source; D. encapsulating the partial image packet The stream is decomposed into partial image signals; E. An integrated image signal is made from the partial image signal, where the integrated image signal includes the codec format; F. An integrated image packet stream is synthesized from the integrated image signal, which includes In the codec format, G. transmit the integrated image packet stream to the first image source. 2. The method according to item 1 of the scope of patent application, further comprising: A. receiving a second connection establishment signal at the network interface, wherein the second connection establishment signal is from a second image source, and Including a second codec identification signal corresponding to the second codec format; B. storing the second codec identification signal in the memory; C. receiving a second portion of the image packet stream by the second image source; D. Decomposing the second part of the image packet stream into the second part of the image signal; E. Using the second part of the image signal to make a second integrated image signal, wherein the second integrated image signal includes the second codec Format; F. synthesize a second integrated video packet stream with the second integrated video signal, which includes the second encoding and decoding format; G. synthesize a third integrated video packet stream with the second integrated video signal, which is

574830 六、申請專利範圍包含該編解碼格式； Η.將該第二集成影像封包流傳送至該第二影像源； I.將該第三集成影像封包傳送至該第一影像源。 3. —種於網路中收發影像流的裝置，其係為一影像處理單元，其包括： (1) 一網路介面，其係用以從一影像源接收建立連線訊號及部分影像封包流並將一集成影像封包流傳送至該影像源，其中該建立連線訊號包含一編解碼格式； (2) —記憶體，其包含一建立連線運算區，其係用以辨認及儲存該建立連線訊號及其編解碼格式； (3) —封包驅動器，其係用以將該部分影像封包流拆解為部分影像訊號，並將一集成影像訊號合成為一集成影像封包流； (4) 一多點控制單元，其係用以將該部分影像訊號變為該集成影像訊號；其中，該封包驅動器係與該多點控制單元連結，該多點控制單元係與該記憶體連結，該網路介面係與該記憶體連結。 4. 如申請專利第3項所述之裝置，其中該網路介面係可以由一第二影像源接收第二建立連線訊號，其中該建立連線訊號包含第二編解碼格式，使得可以於該建立連線運算區中辨識該第二編解碼格式並將之儲存於該記憶體中。 5 .如申請專利第3項所述之裝置，其中該多點控制單元係可順序感測該網路介面是否有來自該第二影像源的第二部574830 6. The scope of patent application includes the codec format; Η. The second integrated image packet stream is transmitted to the second image source; I. The third integrated image packet stream is transmitted to the first image source. 3. —A device for sending and receiving image streams in a network, which is an image processing unit, which includes: (1) a network interface, which is used to receive a connection establishment signal and part of an image packet from an image source Stream and send an integrated image packet stream to the image source, where the connection establishment signal includes a codec format; (2) — memory, which includes a connection establishment operation area, which is used to identify and store the connection Establish the connection signal and its codec format; (3) —packet driver, which is used to decompose part of the image packet stream into part of the image signal, and synthesize an integrated image signal into an integrated image packet stream; (4) ) A multi-point control unit, which is used to change part of the image signal into the integrated image signal; wherein the packet driver is connected to the multi-point control unit, and the multi-point control unit is connected to the memory, the The network interface is connected to the memory. 4. The device as described in claim 3, wherein the network interface can receive a second connection establishment signal from a second image source, and the connection establishment signal includes a second encoding and decoding format, so that the The connection establishment operation area identifies the second codec format and stores it in the memory. 5. The device according to item 3 of the patent application, wherein the multi-point control unit is capable of sequentially sensing whether the network interface has a second part from the second image source.

第34頁 574830 六、申請專利範圍分影像封包流，並在接收該第二部分影像封包流之後，使得該封包驅動器將該第二部分影像封包流拆解為一第二部分影像訊號，且該多點控制單元進而以該編解碼格式及該第二編解碼格式將該第二部分影像訊號變為一第二集成影像訊號，使得該封包驅動器得以該編解碼格式及該第二編解碼格式將該第二集成影像訊號合成為一第二集成影像封包流，使得網路介面能將該編解碼格式之該第二集成影像封包流傳送至該影像源，並能將該第二編解碼格式之該第二集成影像封包流傳送至該第二影像源。Page 34 574830 6. The scope of the patent application is divided into image packet streams, and after receiving the second part of the image packet stream, the packet driver disassembles the second part of the image packet stream into a second part of the image signal, and the The multi-point control unit further changes the second part of the video signal into a second integrated video signal by using the codec format and the second codec format, so that the packet driver can use the codec format and the second codec format. The second integrated image signal is synthesized into a second integrated image packet stream, so that the network interface can transmit the second integrated image packet stream in the encoding and decoding format to the image source, and can transmit the second encoding and decoding format. The second integrated image packet stream is transmitted to the second image source.

6 .如申請專利第3項所述之裝置，其中該網路介面係可以與複數個影像源連結，使得該等影像源能夠傳送符合特定編解碼格是的部分影像封包流至該網路介面，而該網路介面在接收並處理來自該等影像源之該等部分影像封包流之後，能夠依據該等特定的編解碼格式將得出的集成影像封包流分別傳送至該等影像源。6. The device according to item 3 of the patent application, wherein the network interface can be connected to a plurality of image sources, so that the image sources can transmit part of the image packet stream that conforms to a specific codec format to the network interface. After receiving and processing the partial image packet streams from the image sources, the network interface can transmit the resulting integrated image packet streams to the image sources according to the specific codec formats.

第35頁Page 35