TWM304860U - Apparatus with multiple output streams of digital data - Google Patents

Description

M304860 二 八、新型說明： 【新型所屬之技術領域】 本創作係關於-縫位資韻輸出之裝置，特別是關於 一種多重數位資料流輸出之裝置。 【先前技術】 近年來，網路上的資料傳輸，已由早期的文字資料傳輪 發展到目_來越普遍的多媒體網路傳輸。隨著電腦網路與 多媒體的蓬勃發展，影像和視訊編碼麟成為熱Η研究的主 題，影音資訊透過現存的網際網路與無線網路傳送，但因受 到頻寬的限制，使得影像傳輪品質的好壞取決於編碼技術的 優劣。 ί料的壓縮並數位化使影音資料可輕易地在網路上傳 輸，目前市場上之數位影像產品的功能越來越多元且多與網 路結合，如監控系統之網路監控攝影機（Netw〇rk ΙρM304860 Twenty-eight, new description: [New technical field] This creation is about the device for the output of the seam position, especially for a device for outputting multiple digit data streams. [Prior Art] In recent years, data transmission on the Internet has been developed from the early text data transmission to the more popular multimedia network transmission. With the rapid development of computer networks and multimedia, video and video coding has become the subject of ardent research. Video and audio information is transmitted through existing Internet and wireless networks. However, due to the limitation of bandwidth, the quality of image transmission is improved. The quality depends on the pros and cons of the coding technology. The compression and digitization of the material allows the audio and video data to be easily transmitted over the network. Currently, the digital imaging products on the market are more and more diverse and integrated with the network, such as the network monitoring camera of the monitoring system (Netw〇rk). Ιρ

Camera)，數位機上盒(set_t〇p box，STB )、手機甚至是數位 像機等多媒體產品，其中壓縮過之影音資料透過網路傳送到 世界任何一角落，使得享受影音多媒體娛樂變得更加容易。 通常影音資料量都是非常龐大的，而網路之頻寬有限， 因此適當將資料壓縮以利能更快速在網路上傳輸是必備的。 6 M304860 Λ MPEG(Movmg Picture Exports Group)之壓縮標準是由國 際電信聯盟(International Telecommunications Union，ITU)與 國際標準化組織(International Standard Organization，ISO)， 所制定用來壓縮視訊與音訊的標準，並陸續制定出 MPEG]、MPEG-2與MPEG-4等標準，現已成為目前視訊 壓縮的主要技術。MPEG音訊壓縮標準如MPEG-1 layer I(MP1)和 MPEG-1 layer* II(MP2)的壓縮率分別為 4:1 和 6:1-8:1，而MPEG-1 layerIII就是俗稱的MP3壓縮標準，其 壓縮率則高達10:1-12:1。 聲音訊號和影像訊號最大的不同，在於影像是對時間展 開的二維訊號，而聲音則是對時間展開的一維訊號。MPEG 視訊及音訊壓縮的基本架構是先將自然的連續訊號或動作 進行訊號取樣(sampling)、分割為區塊(block)等處理。接著資 料壓縮通常是透過轉換(Transform)和量化(Quantization)過程 來達到此目的。轉換是指將資料由空間域(spatial domain)或 時間域(time domain)轉到頻率域(frequency domain)的過程。 這種轉換，一般是以離散餘弦轉換（Discrete Cosine Transform，DCT)來做轉換。而量化過程在於將高頻訊號給 過濾掉，以達到資料縮減的效果。 JPEG(Joint Photographic Experts Group)是 M304860 ISOdntemational Standardization Orgnizati〇n)以及 CCITT(Consultative Committee Of the Intemation Telephone and Telegraph)所發展出來的靜態連續色調調變的壓縮標 準，JPEG基線(baseline)演算法可以粗略地分成··取樣、離 散餘弦轉換、量化和熵編碼。在；pEG巾，會將編碼好的資 料，取樣分割為數個資料區塊，每—舰塊就稱為最小程式 碼單7C。其中分離餘弦變換，影像由空間域換到頻率域。而 JPEG基線的損耗特性是在量化過程中產生的，藉以剔除影 像中不必要的資訊，量化有助於JpEG的主要壓縮特徵。最 後JPEG蚊了兩種氣碼的方法：霍夫曼編碼和算術編 碼。基線連續數位訊號編解碼器用的是霍夫曼編碼，是一種 針對相片影像而廣泛使㈣―種失錢縮標準方法。 透過上述之mpeg視訊、音訊壓縮標準以及jpEG之壓 縮標準，在訊號經過取樣、切割區塊以及量化等步驟後，無 可避免會與原訊號不同，而為一種失真麼縮方法。 第一圖為一種網路監控攝影機透過網路傳送影音資料之 不意圖。參考第-圖之網路監控攝影機丨⑴，將壓縮過的聲 音及影像播案，透過網路1〇3傳到遠方，使用者可以很方便 使用遠端個人電腦與近端個人電腦、手機浙、個人數 位助理器（personal Digltal八如贫如t，pDA)⑽等設備解壓 M304860 縮影音資料流，即可線上使用瀏覽器瀏覽網路監控攝影機傳 來的影音貧料進行遠端監控。目前市面上己有JPECJ、 MPEGim等細獨壓縮方法之網路監控攝影機，以因應 網路頻寬限制。網路監控攝影機比起傳統·監控攝影機所 具有的一大優勢，也可以將影像資料儲存在儲存設備中如硬 碟 0、網路儲存設備 Attached St〇rage，nas)i ι 卜 網路儲存設備m是一個非常方便使用的網路储存設備，不 但可儲存相當大的資料量並且安裝容易。 網路監控攝影機還有-項功能是允許多個遠端使用者同 時使用職ϋ，劉覽監控畫面。如果某_遠端使用者更改網 路攝影機之解析度(res〇lution)，例如把影像從64_〇大小 改為3施24〇大小，則會造成遠端之所有使用者之解晰度全 部從640x480變更為320爾大小，甚至於儲存於硬碟的影 像解析度也被變更。 目前網路攝影機所使用之壓縮晶片無法針對使用者需求 或是不同的應用’同時輸出不同解析度或是列壓縮比 (compression rati0)之壓縮資料。通常網路儲存設備需要储存 高解析度之影像畫面以利辨識，如64〇χ48〇大小之晝面，而 手機則受限於網路頻寬以及手機營幕大小，只能接收 320x240甚至160x120大小之晝面。 M304860 目月il 1C设计廠商之壓縮晶片的設計均為單一數位資料 流輸出。雖然在資料輸出之前，可以選擇不同之壓縮比以及 不同解析度之壓縮格式，但是每次只能輸出一種壓縮比或解 析度之壓縮資料。 因此’提供一種能夠同時選擇不同之壓縮比以及不同解 析度之壓縮資料格式的裝置，以因應不同使用者之不同需求 或應用有其需要。 第二圖為一種傳統訊號失真編碼系統之示意圖。參考第 二圖，此傳統訊號編碼系統主要可分為四個單元：資料處理 單元201、量化單元203、編碼單元205以及記憶體單元207。 訊號處理單元201主要分析和處理訊號來源並原始訊號轉換 成適合量化的訊號型態，包括一取樣器2〇la、一區塊切割器 201b以及一正交轉換編碼器2〇lc。正交轉換編碼器2〇lc包 含如離散餘弦轉換、離散小波轉換(DiscreteCamera), digital set-top box (set_t〇p box, STB), mobile phone and even digital camera and other multimedia products, in which compressed audio and video data is transmitted to any corner of the world through the network, making video and multimedia entertainment more enjoyable. easily. Usually the amount of audio and video data is very large, and the bandwidth of the network is limited, so it is necessary to properly compress the data to facilitate faster transmission on the network. 6 M304860 Λ MPEG (Movmg Picture Exports Group) compression standard is developed by the International Telecommunications Union (ITU) and the International Organization for Standardization (ISO), which is used to compress video and audio standards. The development of standards such as MPEG], MPEG-2 and MPEG-4 has become the main technology of video compression. MPEG audio compression standards such as MPEG-1 layer I (MP1) and MPEG-1 layer* II (MP2) have compression ratios of 4:1 and 6:1-8:1, respectively, while MPEG-1 layer III is commonly known as MP3 compression. The standard has a compression ratio of 10:1-12:1. The biggest difference between an audio signal and an image signal is that the image is a two-dimensional signal that is spread over time, and the sound is a one-dimensional signal that is developed over time. The basic structure of MPEG video and audio compression is to first process and divide the natural continuous signal or motion into blocks. Subsequent data compression is usually achieved through the transformation and quantization processes. Conversion refers to the process of transferring data from a spatial domain or a time domain to a frequency domain. This conversion is usually done by Discrete Cosine Transform (DCT). The quantification process is to filter out the high-frequency signals to achieve the effect of data reduction. JPEG (Joint Photographic Experts Group) is a compression standard for static continuous tone modulation developed by M304860 ISOdntemational Standardization Orgnizati〇n) and CCITT (Consultative Committee Of the Intemation Telephone and Telegraph). The JPEG baseline algorithm can be roughly used. Divided into · sampling, discrete cosine transform, quantization and entropy coding. In the ;pEG towel, the coded information is divided into several data blocks, and each block is called the minimum code list 7C. Where the cosine transform is separated, the image is switched from the spatial domain to the frequency domain. The loss characteristics of the JPEG baseline are generated during the quantization process, in order to eliminate unnecessary information in the image, and the quantization contributes to the main compression features of JpEG. Finally, JPEG mosquitoes have two methods of air code: Huffman coding and arithmetic coding. The baseline continuous digital codec uses Huffman coding, which is a standard method for the photographic image to make (4) a variety of money. Through the above mpeg video, audio compression standard and jpEG compression standard, after the signal is sampled, cut and quantized, it is inevitable that it will be different from the original signal, and it is a distortion method. The first picture shows a network surveillance camera that transmits video and audio data over the network. Refer to the network monitor camera 第(1) in the figure, and transmit the compressed sound and video to the remote through the network 1〇3. The user can conveniently use the remote PC and the near-end personal computer and mobile phone. The personal digital assistant (personal Digltal eight, such as poor t, pDA) (10) and other devices decompress the M304860 video stream, you can use the browser to browse the video surveillance camera from the network to remote monitoring. At present, there are network surveillance cameras with fine compression methods such as JPECJ and MPEGim on the market, in order to respond to network bandwidth restrictions. Compared with traditional surveillance cameras, network surveillance cameras can also store image data in storage devices such as hard disk 0, network storage device Attached St〇rage, nas) i ι 卜 network storage device m is a very convenient network storage device, which can store a large amount of data and is easy to install. The network surveillance camera also has a function that allows multiple remote users to use the job at the same time. If a _ remote user changes the resolution of the webcam, for example, changing the image size from 64_〇 to 3 〇24〇, it will cause the resolution of all users at the far end. Changed from 640x480 to 320 s, and even the image resolution stored on the hard drive has been changed. At present, the compressed chips used in the network camera cannot output compressed data of different resolutions or compression ratios at the same time for the user's needs or different applications. Usually, network storage devices need to store high-resolution image images for identification, such as 64〇χ48〇 size, while mobile phones are limited by network bandwidth and mobile screen size, and can only receive 320x240 or even 160x120 size. After the face. M304860 The design of the compressed chip of the il 1C design manufacturer is a single digital data stream output. Although different compression ratios and compression formats with different resolutions can be selected before the data is output, only one compression ratio or resolution compression data can be output at a time. Therefore, a device capable of simultaneously selecting compressed data formats of different compression ratios and different resolutions is provided to meet the needs of different users or applications. The second figure is a schematic diagram of a conventional signal distortion coding system. Referring to the second figure, the conventional signal encoding system can be mainly divided into four units: a data processing unit 201, a quantization unit 203, an encoding unit 205, and a memory unit 207. The signal processing unit 201 mainly analyzes and processes the signal source and converts the original signal into a signal type suitable for quantization, including a sampler 2〇1a, a block cutter 201b, and an orthogonal transform encoder 2〇lc. The orthogonal transform coder 2〇lc includes, for example, discrete cosine transform, discrete wavelet transform (Discrete)

WaveletWavelet

Transform ’ DWT)、KLT(Karhunen-Loeve Transform)轉換 等’目的是把在空間域之訊號變換到轉換域(如頻率域），並 以轉換係數來表示。 里化單元203包含一量化器2〇3a與一量化表(quantization M304860 - table)203b。量化表2〇3b裡之元素值比較小時，壓縮比會比 較低，相對地影像品質比較好。相反地，當量化表2〇3b内 的元素值偏大時，則壓縮比會比較高，但影像品質較差。 編石馬單元205接收量化單元203輸出之資料，以進行編 碼、壓縮，並將壓縮過後之資料傳至記憶體單元2〇7内暫存 然後輸出。 中華民國專利第545,827號揭露一種可降低數位顯示器 數位影像資料傳輸資料量的架構，其具有-數姆彡像及時序 控制單元，用來控制數位影像資料之接收或傳送。 中華民國專利第39〇,〇97號揭露一種可選擇壓縮比的影 像壓縮系統，此系統係透過一種檢測影像複雜度的方法，來 調i JPEG中里化器的置化間距，因而維持影像晝質與調整 壓縮比。 申華民國專利第566,047號揭露一種多重速率數位資料 流的代碼轉換機，從一制之輸出位元資料流，產生多個不 同速率之多重重新編碼的位元資料流。 M304860 【新型内容】 鑑於上述前述之傳紐位龍流之單—輸出問題。因此 本創作提供-種多重數位資料流輸出裝置，此裝置提供多個 不同之數位資料流同時輪出。 本裝置處理包含聲音以及景緣資料之訊號資料，經過資 料處理、以及量化編碼等步驟後，無可避免會與原訊號不 同，所以是一種失真壓縮裝置。The purpose of Transform ' DWT), KLT (Karhunen-Loeve Transform) conversion, etc. is to transform the signal in the spatial domain into a conversion domain (such as a frequency domain) and express it by a conversion coefficient. The refinement unit 203 includes a quantizer 2〇3a and a quantization table 304 (quantization M304860-table) 203b. When the element values in the quantization table 2〇3b are small, the compression ratio will be lower, and the image quality is better. Conversely, when the value of the element in the quantization table 2〇3b is too large, the compression ratio is relatively high, but the image quality is poor. The stone unit 205 receives the data output by the quantization unit 203 for encoding and compression, and transfers the compressed data to the memory unit 2〇7 for temporary storage and then output. The Republic of China Patent No. 545,827 discloses an architecture for reducing the amount of digital image data transmitted by a digital display, which has a digital image and timing control unit for controlling the reception or transmission of digital image data. The Republic of China Patent No. 39, 〇97 discloses an image compression system with a selectable compression ratio. This system adjusts the localization spacing of the JPEG in the JPEG by means of a method for detecting image complexity, thereby maintaining image quality. Adjust the compression ratio. Shenhua, Republic of China Patent No. 566,047 discloses a transcoding machine for multi-rate digital data streams, which generates a plurality of multi-recoded bit data streams of different rates from a single output bit stream. M304860 [New content] In view of the above-mentioned single-output problem of the new line. Therefore, this creation provides a multi-digit data stream output device that provides a plurality of different digital data streams to be rotated simultaneously. The device processes the signal data including the sound and the edge data. After the steps of data processing and quantization coding, it is inevitably different from the original signal, so it is a distortion compression device.

I 本裝置包含-資料處理單元、_選擇器單元以及一量化 編碼單π，資料褚理單元產生各種不同解析度之訊號資料。 選擇器單70將各種不同解析度之訊號資料分配組合輸出至 量化編碼單元，即可依·化表内之元素數值產生各種不同 壓縮比之訊號資料。 本裝置更備有足夠之記憶體以儲存數位資料，並依照需 求輪出適當之數位資料。 k著半導體產業的進步，在—單晶片裡可放的晶元越來 越多，而尺寸則越來越小。因此，本創作之多重數位資料流 輪出裝置也可實現於一單晶片上。 M304860 茲配合下列圖示、實施例之詳細說明及申請專利範圍， 將上述及本創作之其他目的與優點詳述於後。 【實施方式】 第二圖為本創作之多重數位資料流輸出裝置的一個示音 圖。參考第三圖，此多重數位資料流輸出裝置包含一資料處 理單元301、一選擇器單元303、一量化編碼單元3〇5、一記 憶體單元307以及一輸出單元309。 資料處理單元301内備有一個或一個以上之取樣器 301a、區塊切割器3〇lb以及正交轉換編碼器301c，處理輸入 之訊號資料並產生多個取樣轉換資料，其中多個取樣轉換資 料更包含了多個不同解析度之資料。 選擇器單元303接收資料處理單元301輸出之多個取樣 轉換資料，並且根據輸出單元3〇9之資料分配資訊，將此多 個轉換資料傳送至量化編碼單元305。量化編碼單元3〇5内 包括多個量化表305a、量化器305b以及編碼器3〇5c。將包 括選擇器單元303輸出之多個取樣轉換資料作量化和編碼等 處理，並產生多個量化編碼資料。此多個量化編碼資料包括 多個不同解析度以及壓縮比之數位資料。 13 M304860 $憶體單元斯具有足夠大小之記憶體以暫存多個量化 編喝資料輸入。輪出單元彻侧網路流量、使用者需求等 =數，讀取記憶體單元307内之量化編碼資料。其量化編碼 貝料即包括不同解析度以及壓縮比之數位資料。 • 帛四®為本創作第—實施例之多重數位資料流輸出裝 置。參考第四圖，此裝置包含一資料處理單元姻、一選擇 • 态單元303、—量化編碼單元405、一記憶體單元307以及一 輸出單元309。 資料處理單元401包括一取樣器401a、一區塊切割器 401b以及一離散餘弦轉換器4〇lc，將輸入之影像資料作取樣 以及轉換後產生一 640x480解析度之取樣轉換資料並輸出至 選擇器單元303。選擇器單元303係一旁路(bypass)選擇器， 即此選擇器單元303接收由輸出單元309輸出之資料分配資 訊後將640X480解析度之取樣轉換資料，直接輸出至量化編 碼單元405。 量化編碼單元405包括三個量化器405d，405e，405f以及 三個嫡編碼器405g，405h，405i以及用來決定資料壓縮比的量 化表405a，405b，405c。量化表405a内所儲存的元素值為三個 量化表(405a，405b，405c)中之最高，因此被壓縮後之影像資料 14 M304860 量為最小。量化表405c内所儲存的元素值為三個量化表中之 最低’因此被壓縮後之影像資料量最大。量化表405b内所儲 存的元素值為介於量化表405a以及405c中間，因此被壓縮 後之影像資料量介於中間。 換言之，一影像資料由資料處理單元401輸出之640x480 解析度之取樣轉換資料，經由選擇器單元303直接傳送至量 化編碼單元405作量化編碼處理後，輸出三個640x480解析 度以及不同壓縮比之量化編碼資料並暫存至記憶體單元3〇7 中。其中包括一 640x480低壓縮比之影像資料4〇7a、一 640x480中壓縮比之影像資料4〇7b以及一 640x480高壓縮比 之影像資料407c。 最後經由輸出單元309偵測遠端使用者之網路頻寬以及 需求讀取§己憶體單元307之量化編碼資料並傳輸至網路上。 其中輸出單元309可傳送一 640x480高壓縮比之影像資料 407c至個人電腦端以及一 640x480中壓縮比之影像資料407b 至個人數位助理器，以達同時輸出不同之數位訊號至使用者 端0 當遠端之使用者如個人電腦，若受限於網路頻寬而只能 接收6输柳南壓縮比之影像資料屻^則個人電腦端更可 15 M304860 4 以將影像> 料作後處理以改善觀看之品質。即在解麽，缩 640x480南壓縮比之影像資料4〇7c時，將解析度縮小 (downscaling)為320x240甚至於160xl2〇大小，降低影像厶 顆粒狀，更進而改善影像觀看之品質。 第五圖為本創作之多重數位資料流輸出裝置的一第二實 施例。參考第五圖，此多重數位資料流輸出裝置包含一資料 處理單元501(内含三組取樣器、區塊切割器、離散餘弦轉換 •器）、一選擇器單元303、一量化編碼單元5〇5(内含三組量化 器以及量化表）、一記憶體單元307以及一輸出單元309。 第五圖中一影像資料係一 640x480之原始影像，傳進訊 號處理單元501並分別產生640x480、320x240、160x120三 種解析度之影像501a-501c。選擇器單元303接收輸出單元 309之資料分配資訊，其中網路硬碟以及遠端個人電腦要求 B 640x480低壓縮比之影像資料。手機要求160xl2〇高壓縮比 之影像資料。個人數位助理器要求320x240中壓縮比之影像 資料。因此，選擇器單元303將640x480、320x240、160x120 三種不同解析度之影像501a-501c組合並分配至量化編碼單 元505作量化、編碼處理。 量化編碼單元505内含三組量化表505a - 505c、三個量 16 M304860 」 化器505d- 505f和三個嫡編碼器505g- 505i。量化表5〇5a内 所儲存的元素值為三個量化表中之最高’因此影像資料之壓 縮比最高。量化表505c内所儲存的元素值為三個量化表中之 最低，因此影像資料之壓縮比最低。量化表505b内所儲存的 、 元素值為介於量化表505a以及505c中間，因此影像資料之 壓縮比介於中間。 g 換言之，一 640x480之原始影像經由資料處理單元5〇卜 選擇器單元303以及量化編碼單元505後，輸出三個量化編 碼資料暫存於記憶體單元307中，此三個量化編碼資料包括 一 640x480低壓縮比之影像資料507a、一 320x240中壓縮比 之影像資料507b以及一 160x120高壓縮比之影像資料5〇7c。 最後經由輸出單元309讀取記憶體單元内307所暫存之 φ 影像資料經由網路傳輸至使用者端。其_輸出單元309傳送 160x120高壓縮比之影像資料5〇7c至手機端；64〇χ48〇低壓 縮比之影像資料507a至網路硬碟以及個人電腦端；32〇χ24〇 中壓縮比之影像資料5〇7b至個人數位助理器端，以達同時輸 出不同之數位訊號至使用者端。. 惟，以上所述者，僅為本創作之列舉實施例而已，當不 如此限林創作實叙顧。即大凡依本創作_請專利範 17 M304860 圍所作之均等變化與修飾，皆應仍屬本創作專利涵蓋之範圍 S ^ 内0I The device comprises a data processing unit, a _selector unit and a quantized coding unit π, and the data processing unit generates signal data of various resolutions. The selector unit 70 outputs the signal data combination combinations of various resolutions to the quantization coding unit, and can generate signal data of various compression ratios according to the element values in the table. The device is also equipped with sufficient memory to store digital data and to rotate the appropriate digital data as required. With the advancement of the semiconductor industry, more and more wafers can be placed in a single wafer, and the size is getting smaller and smaller. Therefore, the multi-digit data streamer of the present invention can also be implemented on a single wafer. The above and other objects and advantages of the present invention will be described in detail below with reference to the following drawings, detailed description of the embodiments and claims. [Embodiment] The second figure is a sound map of the multi-digit data stream output device of the present invention. Referring to the third figure, the multi-bit data stream output device comprises a data processing unit 301, a selector unit 303, a quantized coding unit 3〇5, a memory unit 307, and an output unit 309. The data processing unit 301 is provided with one or more samplers 301a, a block cutter 3〇b, and an orthogonal transform encoder 301c for processing the input signal data and generating a plurality of sample conversion data, wherein the plurality of sample conversion data It also contains a number of different resolution data. The selector unit 303 receives the plurality of sample conversion data outputted by the data processing unit 301, and transmits the plurality of conversion data to the quantization coding unit 305 according to the data distribution information of the output unit 3〇9. The quantization coding unit 3〇5 includes a plurality of quantization tables 305a, quantizers 305b, and encoders 3〇5c. The plurality of sample conversion data outputted by the selector unit 303 is subjected to processing such as quantization and encoding, and a plurality of quantized coded data are generated. The plurality of quantized encoded data includes a plurality of digital data of different resolutions and compression ratios. 13 M304860 $Restore unit has enough memory to temporarily store multiple quantified data. The round-out unit completes the network traffic, user demand, etc., and reads the quantized coded data in the memory unit 307. The quantized coding material includes digital data of different resolutions and compression ratios. • ®4® is the multi-digit data stream output device of the author's first embodiment. Referring to the fourth figure, the apparatus includes a data processing unit, a selection unit 303, a quantization coding unit 405, a memory unit 307, and an output unit 309. The data processing unit 401 includes a sampler 401a, a block cutter 401b, and a discrete cosine converter 4〇lc. The input image data is sampled and converted to generate a 640×480 resolution sample conversion data and output to the selector. Unit 303. The selector unit 303 is a bypass selector. That is, the selector unit 303 receives the data distribution information outputted by the output unit 309 and outputs the sample conversion data of 640×480 resolution to the quantization and coding unit 405. The quantized coding unit 405 includes three quantizers 405d, 405e, 405f and three chirp encoders 405g, 405h, 405i and quantization tables 405a, 405b, 405c for determining the data compression ratio. The element value stored in the quantization table 405a is the highest of the three quantization tables (405a, 405b, 405c), so the amount of the compressed image data 14 M304860 is the smallest. The value of the element stored in the quantization table 405c is the lowest of the three quantization tables, and thus the amount of image data after compression is the largest. The value of the element stored in the quantization table 405b is intermediate between the quantization tables 405a and 405c, so that the amount of compressed image data is intermediate. In other words, the 640x480 resolution sample conversion data outputted by the data processing unit 401 by the data processing unit 401 is directly transmitted to the quantization coding unit 405 via the selector unit 303 for quantization and coding processing, and then outputs three 640x480 resolutions and different compression ratio quantization. The encoded data is temporarily stored in the memory unit 3〇7. These include a 640x480 low compression ratio image data 4〇7a, a 640x480 compression ratio image data 4〇7b, and a 640x480 high compression ratio image data 407c. Finally, the output unit 309 detects the network bandwidth of the remote user and needs to read the quantized coded data of the hex element 307 and transmit it to the network. The output unit 309 can transmit a 640x480 high compression ratio image data 407c to the personal computer end and a 640x480 compression ratio image data 407b to the personal digital assistant to simultaneously output different digital signals to the user terminal. The user of the end, such as a personal computer, can only receive 6 images of Liunan compression ratio if it is limited by the network bandwidth. Then the PC can be 15 M304860 4 to post-process the image > Improve the quality of viewing. That is to say, when the solution is reduced to 640x480 and the image data is 4〇7c, the resolution is reduced to 320x240 or even 160xl2〇, which reduces the image granules and improves the quality of image viewing. The fifth figure is a second embodiment of the multi-digit data stream output device of the present invention. Referring to the fifth figure, the multi-digit data stream output device comprises a data processing unit 501 (containing three sets of samplers, block cutters, discrete cosine transformers), a selector unit 303, and a quantized coding unit. 5 (containing three sets of quantizers and quantization tables), a memory unit 307, and an output unit 309. In the fifth figure, an image data is a 640x480 original image, which is transmitted to the signal processing unit 501 and generates images 501a-501c of three resolutions of 640x480, 320x240, and 160x120, respectively. The selector unit 303 receives the data distribution information of the output unit 309, wherein the network hard disk and the remote personal computer require B 640x480 low compression ratio image data. The mobile phone requires 160xl2〇 high compression ratio image data. The personal digital assistant requires a compression ratio of 320x240. Therefore, the selector unit 303 combines and distributes the three different resolution images 501a-501c of 640x480, 320x240, and 160x120 to the quantization encoding unit 505 for quantization and encoding processing. The quantization coding unit 505 contains three sets of quantization tables 505a - 505c, three quantities 16 M304860" 505d - 505f and three 嫡 encoders 505g - 505i. The element values stored in the quantization table 5〇5a are the highest among the three quantization tables', so the compression ratio of the image data is the highest. The element values stored in the quantization table 505c are the lowest of the three quantization tables, so the compression ratio of the image data is the lowest. The element values stored in the quantization table 505b are intermediate between the quantization tables 505a and 505c, so that the compression ratio of the image data is intermediate. In other words, a raw image of 640x480 is outputted to the memory unit 307 by the data processing unit 5, the selector unit 303 and the quantization and encoding unit 505, and the three quantized encoded data includes a 640x480. Low compression ratio image data 507a, a 320x240 compression ratio image data 507b, and a 160x120 high compression ratio image data 5〇7c. Finally, the φ image data temporarily stored in the memory unit 307 is read via the output unit 309 and transmitted to the user end via the network. The output unit 309 transmits 160×120 high compression ratio image data 5〇7c to the mobile phone terminal; 64〇χ48〇 low compression ratio image data 507a to the network hard disk and the personal computer terminal; 32〇χ24〇 compression ratio image The data 5〇7b to the personal digital assistant end, so as to simultaneously output different digital signals to the user end. However, the above-mentioned ones are only examples of the examples of this creation, and it is not so limited to the creation of the forest. That is to say, the equivalent change and modification made by the patent Scope 17 M304860 should still be within the scope of this creation patent.

18 M304860 【圖式簡單說明】 - 第一圖為說明網路監控攝影機透過網路傳送資料之示 意圖。 第二圖為傳統訊號編碼系統示意圖。 第三圖為本創作之多重數位資料流輸出裝置之示 • 意圖。 第四圖為本創作之多重數位資料流輸出裝置的第 一實施例。 第五圖為本創作之多重數位資料流輸出裝置的第 【主要元件符號說明】 101網路監控攝影機 103網路 105個人電腦 107手機 108個人數位助理 110硬碟 111網路儲存設備 201資料處理單元 201a取樣器 201b區塊切割器 201c正交轉換編碼器 203量化編碼單元 203a量化器 203b量化表 1918 M304860 [Simple description of the diagram] - The first figure shows the intention of the network surveillance camera to transmit data through the network. The second picture is a schematic diagram of a conventional signal coding system. The third figure is the intent of the multi-digit data stream output device of the creation. The fourth figure is a first embodiment of the multi-digit data stream output device of the present invention. The fifth figure is the first part of the multi-digit data stream output device of the creation. [Main component symbol description] 101 network surveillance camera 103 network 105 personal computer 107 mobile phone 108 personal digital assistant 110 hard disk 111 network storage device 201 data processing unit 201a sampler 201b block cutter 201c orthogonal transform encoder 203 quantization coding unit 203a quantizer 203b quantization table 19

M30486H 203c編碼器 205編碼單元 207記憶體單元 301資料處理單元 301a取樣器 301b區塊切割器 301c正交轉換編碼器 303選擇器單元 305量化編碼單元 305a量化器 305b量化表 305c編碼器 307記憶體單元 309輸出單元 401資料處理單元 401a取樣器 401b區塊切割器 401c正交轉換編碼器 405量化編碼單元 405a，405b，405c 量化表 405d，405e，405f 量化器 405g，405h，405i 編碼器 407a 640x480低壓縮比之影 像資料 407b 640x480中壓縮比之影 像資料 407c 640x480高壓縮比之影 像資料 501資料處理單元 505量化編碼單元 20 505a，505b，505c 量化表 505d，505e，505f 量化器 505g，405h，505i 編碼器 507a 640x480低壓縮比之影 507b 320x240中壓縮比之影 像資料 像資料 507c 160x120高壓縮比之影 像資料M30486H 203c encoder 205 encoding unit 207 memory unit 301 data processing unit 301a sampler 301b block cutter 301c orthogonal transform encoder 303 selector unit 305 quantization coding unit 305a quantizer 305b quantization table 305c encoder 307 memory unit 309 output unit 401 data processing unit 401a sampler 401b block cutter 401c orthogonal transform encoder 405 quantization coding unit 405a, 405b, 405c quantization table 405d, 405e, 405f quantizer 405g, 405h, 405i encoder 407a 640x480 low compression More than the image data 407b 640x480 compression ratio image data 407c 640x480 high compression ratio image data 501 data processing unit 505 quantization coding unit 20 505a, 505b, 505c quantization table 505d, 505e, 505f quantizer 505g, 405h, 505i encoder 507a 640x480 low compression ratio shadow 507b 320x240 compression ratio image data 507c 160x120 high compression ratio image data

M304&6QM304&6Q

21twenty one

Claims (1)

M304860 九、申請專利範圍： ‘ 1· 一種多重數位資料流輸出裝置，該裝置包含： 一資料處理單元，接收一輸入資料並輸出多個取樣轉換資 料； 一選擇器單元，接收輸入一資料分配以及該多個取樣轉換資 • 料，並分配輸出該多個取樣轉換資料； 一量化編碼單元，接收該多個取樣轉換資料，並輸出多個量 化編媽資料； 一記憶體單元，接收該多個量化編碼資料暫存；以及 T雨出單元’輸出該資料分配資訊，並讀取該記憶體單元之 該量化編碼資料後輸出。 2·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中該資料處理單元内備有兩個或兩個以上之取樣器、區塊切 割器以及正交轉換編碼器。 3.如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中該選擇H單元係-旁路選擇H，即該選顧將接收之該取 _ 樣轉換資料直接輸出至該量化編碼單元。 4·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中該選擇器單元將接收之該兩個或兩個以上之取樣轉換資 料組合分配後輸出至該量化編碼單元。 5·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中該i化編碼單元中備有至少兩個或兩個以上之量化器、量 化表以及編碼器。 6·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中兩個或兩個以上之該取樣轉換資料，係包括兩個或兩個以 22 M304860 .一 上解析度之資料。 . 7·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 中兩個或兩個以上之該量化編碼資料，係包括兩個或兩個以 上解析度以及壓縮比之資料。 8·如申請專利範圍第1項所述之多重數位資料流輸出裝置，其 ’中該圮憶體單元係被備有足夠之記憶體以儲存該量化編碼 資料。 9· 一種多重數位資料流輸出裝置，該裝置包含·· _ 一資料處理單元，接收一輸入資料並輸出一取樣轉換資料； 選擇器單元，接收輸入一資料分配以及該取樣轉換資料， 並分配輸出該多個取樣轉換資料； 一量化編碼單元，接收該多個取樣轉換資料，並輸出多個量 化編碼資料； 一記憶體單元，接收該多個量化編碼資料暫存；以及 -輸出單元，輸出該賴分§&資訊，並讀取該記憶體單元之 該量化編碼資料後輸出。 _ 10.如申明專利I巳圍第9項所述之多重數位資料流輸出裝置，其 中該資料處理單元内包括-取樣器、—區塊蝴器以及一正 交轉換編碼器。 11·如申請專利範圍第9項所述之多重數位資料流輸出裝置，其 找聰H單元係-旁路聰H，即該·雜接收之該取 樣轉換ί料直接輸出至該量化編碼單元。 12·如申#專利範圍第9項所述之多重數位資料流輸出裝置，其 中該量化編碼單元中備有至少兩個或兩個以上之量化器、量 化表以及編碼器。 23 M304860 \ 13.如申請專利範圍第9項所述之多重數位資料流輸出裝置其 中該兩個或兩個以上之量化編碼資料，係兩個或兩個以上解 析度以及壓縮比之資料。 14. 如申請專利範圍第9項所述之多重數位資料流輸出裝置，其 中該記憶體單元係被備有足夠之記憶體以儲存兮 資料。 量化編碼M304860 IX. Patent application scope: '1· A multi-digit data stream output device, the device comprises: a data processing unit, receiving an input data and outputting a plurality of sample conversion data; a selector unit, receiving input and a data distribution and The plurality of samples are converted into materials and distributed to output the plurality of sampled conversion data; a quantized coding unit receives the plurality of sampled conversion data and outputs a plurality of quantized mother data; and a memory unit receives the plurality of The quantized coded data is temporarily stored; and the T rain-out unit outputs the data distribution information, and reads the quantized coded data of the memory unit and outputs the same. 2. The multi-digit data stream output device of claim 1, wherein the data processing unit is provided with two or more samplers, block cutters, and orthogonal transform encoders. 3. The multi-digit data stream output device according to claim 1, wherein the selection H unit is a bypass selection H, that is, the selection directly outputs the received conversion data to the quantization code. unit. 4. The multi-digit data stream output device of claim 1, wherein the selector unit allocates the received two or more sample conversion data combinations to the quantized coding unit. 5. The multi-digit data stream output device of claim 1, wherein the i-coding unit has at least two or more quantizers, quantization tables, and encoders. 6. The multi-digit data stream output device according to claim 1, wherein the two or more of the sample conversion data include two or two data of 22 M304860. 7. The multi-digit data stream output device according to claim 1, wherein the two or more of the quantized coded data include data of two or more resolutions and compression ratios. 8. The multi-digit data stream output device of claim 1, wherein the memory unit is provided with sufficient memory to store the quantized coded data. 9. A multi-digit data stream output device, comprising: a data processing unit, receiving an input data and outputting a sample conversion data; a selector unit receiving input data distribution and the sample conversion data, and distributing the output The plurality of sample conversion data; a quantization coding unit, receiving the plurality of sample conversion data, and outputting a plurality of quantized coded data; a memory unit receiving the plurality of quantized coded data temporarily stored; and an output unit outputting the The § & information is read, and the quantized coded data of the memory unit is read and output. 10. The multi-digit data stream output device of claim 9, wherein the data processing unit comprises a -sampler, a block butterfly, and a quadrature conversion encoder. 11. The multi-digit data stream output device according to claim 9 of the patent application, wherein the C-H unit is bypassed, that is, the sample conversion is directly output to the quantized coding unit. 12. The multi-digit data stream output device of claim 9, wherein the quantized coding unit includes at least two or more quantizers, quantization tables, and encoders. 23 M304860 \ 13. The multi-digit data stream output device of claim 9, wherein the two or more quantized coded data are data of two or more degrees of resolution and compression ratio. 14. The multi-digit data stream output device of claim 9, wherein the memory unit is provided with sufficient memory to store the data. Quantization coding 24twenty four