4. 4 4·4 4 3 6 經 濟 部 財 產 局 員 工 消 費 合 作 社 印 製 Α7 __—____Β7____ 五、發明說明（1 ) 本申請案是於1998年7月1曰提出申請而標題為，，一 用以改善類比至數位轉換器解析度之後置轉換系統，，的申 請案之一部分後續案。 類比至數位轉換器（ADC)將一類比電壓轉換成一以數 5位位元形式表示之二進位表示型態。該ADC的解析度係 以類比電壓被轉換成之數位位元的數目來表示；通常，所 希望的解析度越高’則在該類比至數位轉換器之元件中所 須之精密度越高；例如，放大電晶體等等需要具有更緊密 的容許誤差並且要非常精確，以能得到一具有意義之最小 10 有效位元解析度。 在成像應用上，則希望能改善影像品質，而用以實現 此一願望的一個方法就是在一子系統的輸出端使用一升壓 遽波器，然而’因為升壓濾波器也會放大雜訊，故它們典 型地被使用在尚對比的情況，而非用在具有一低訊號雜訊 15 比的應用中。 本發明認知到對於視覺系統’一接收器（例如人類的 眼睛）只有在某些空間頻率範圍是最敏感的；改善ADC訊 號雜訊比的一個頻率範圍被安排成和在該接收器中最高靈 敏度之頻率範圍相配合。於是，高解析度和準確度之一類 | | 20比至數位轉換只提供給眼睛對其最靈敏的這樣一個頻率範 圍’此一範圍外面的頻率則會以較低解析度和準確度被轉 成數位形式；這導致訊號雜訊比（SNR)會針對所注重的頻 譜而被改善。因此，只須針對該頻譜的一部分作一高準確 度轉換，即可有效地增加影像品質之感知力。 ______第4頁 本紙張尺度適时關家標準（CNS)A(1規格⑵Q χ ------ {請先閱讀背面之沒意事項再填寫本頁) ^ ill —---^---------产 V 4 4443 6 A7 B7 經 濟 部 智 慧 財 產 局 員 工 消 費 合 作 社 印 製 五、發明說明（2 ) 本發明在一實施例中係改善在一系統之輸入端的訊號 雜訊比’而非輸出端；對應於一被感測到之影像的類比訊 號被產生’頻譜將會隨因該影像之不同區域而變化。和其 他類比訊號比較’對於那些類比訊號的空間頻率將會改 變；所有的類比訊號都被轉換成數位訊號來處理，但那些 在所注重的空間頻率範圍中的類比訊號則以較高解析度和 準碟度被轉換。埠降低了在所注重的範圍中的雜訊最低準 位。該頻率可以依據與在一水平線、一垂直線或一區域的 其他樣本之關係而被建立。 在一實施例中，該高準確度轉換之頻率範圍可針對不 同顏色而改變。或者，某些顏色會具有高準確度之轉換， 而其他顏色皆為較低準確度之轉換。 在一實施例中，一類比濾波器被用來選取將會接受高 準確度轉換之資料。在另一實施例中，對來自一 ADC之 已轉換數位位元做後處理，用來提供一個或更多個附加位 元之解析度；例如，一數位濾波器被用來由來自該ADc 的輸出端之鄰近樣本插入額外的位元。 為了更進-步了解本發明之特質和優點，則應參照以 下配合附圖所作之描述。 第1A圖係^示-個ADC的習知技術頻错響應之 5 10 15 20 圖面 續'圖表 第1B圖係根據本發明之頻譜響應。 第2A圖係-繪示-調變臨界值對空間頻率的對數標 第5頁4. 4 4 · 4 4 3 6 Printed by the Consumer Cooperatives of the Property Bureau of the Ministry of Economic Affairs Α7 _______ Β7 ____ V. Description of Invention (1) This application was filed on July 1, 1998 and the title is, one for Post-conversion system to improve analog-to-digital converter resolution, part of the application. An analog-to-digital converter (ADC) converts an analog voltage into a binary representation in the form of a 5-bit number. The resolution of the ADC is expressed by the number of digital bits into which the analog voltage is converted; usually, the higher the desired resolution, the higher the precision required in the analog-to-digital converter component; For example, amplifying transistors and the like need to have tighter tolerances and be very accurate to obtain a meaningful minimum 10 significant bit resolution. In imaging applications, I hope to improve the image quality, and one way to achieve this is to use a booster at the output of the subsystem, but 'because the booster filter also amplifies noise Therefore, they are typically used in contrasting situations rather than in applications with a low signal-to-noise ratio of 15%. The present invention recognizes that for a visual system 'a receiver (such as the human eye) is most sensitive only in certain spatial frequency ranges; a frequency range that improves the ADC signal-to-noise ratio is arranged to have the highest sensitivity in the receiver. The frequency range matches. Therefore, a class of high resolution and accuracy | | 20 ratio to digital conversion is only provided to the frequency range to which the eye is most sensitive. 'Frequency outside this range will be converted with lower resolution and accuracy. Digital form; this results in an improved signal-to-noise ratio (SNR) for the spectrum of interest. Therefore, only a high-accuracy conversion is required for a part of the frequency spectrum to effectively increase the perception of image quality. ______Page 4 This paper's standard timely home care standard (CNS) A (1 size ⑵Q χ ------ {Please read the unintentional matter on the back before filling out this page) ^ ill ------- ^- -------- Production V 4 4443 6 A7 B7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs V. Description of the invention (2) In one embodiment, the present invention improves signal noise at the input end of a system The ratio is not the output; the analog signal corresponding to a sensed image is generated. The frequency spectrum will vary with different areas of the image. Comparing with other analog signals' The spatial frequencies of those analog signals will change; all analog signals are converted to digital signals for processing, but those in the spatial frequency range of interest focus on higher resolution and The accuracy is converted. The port reduces the minimum level of noise in the range of interest. The frequency may be established in relation to other samples in a horizontal line, a vertical line, or an area. In one embodiment, the frequency range of the high-accuracy conversion can be changed for different colors. Alternatively, some colors will have high-accuracy conversions, while others will be lower-accuracy conversions. In one embodiment, an analog filter is used to select data that will undergo high-accuracy conversion. In another embodiment, the converted digital bits from an ADC are post-processed to provide a resolution of one or more additional bits; for example, a digital filter is used to convert the digital bits from the ADc Adjacent samples at the output end insert extra bits. In order to further understand the characteristics and advantages of the present invention, reference should be made to the following description in conjunction with the accompanying drawings. Figure 1A shows a conventional frequency error response of an ADC. 5 10 15 20 Figures Continued 'Figure Figure 1B shows the spectral response according to the present invention. Figure 2A-Plotting-Logarithmic scale of modulation threshold vs. spatial frequency Page 5
— Ill — 11 — II — ^ ^-------i^ilnl —-- <請先閱讀背面之注意事項再填寫本頁) 444 ^3 6 A7 B7 經 濟 部 智 慧 財 產 局 員 工 消 費 合 作 社 印 製 五、發明說明（3) 一 — ' 第2B圖係一繪示調變轉換函數對空間頻率的圖表 第3圖係本發明之一範例實施例的方塊圓。 第4圖係一繪示本發明之一實施例的線性插補法和臨 界值的圖表。 第5圖係一繪不本發明的方法之一實施例的流程圖 第6圖係根據本發明用以提供二個附加位元的解析度 的一個電路之一實施例的方塊圖。 、第7圖係本發明用於一單一色彩或黑白明視度之一模 式的方塊圖。 第8圖係本發明用於多重色彩或不同空間頻率之第二 模式的方塊圖。 第1A圖以一曲線i 〇〇繪示該類比至數位轉換器 之訊號雜訊比（蠢)對頻率⑴之值。由圖可見，該歐 大勺保持於66，直到非常高的頻率為止，即在該範圍内 本質上保持恒定。 第圖繪不該第1A圖被用於本發明之圖表。如同 所見，該SNR開始時在部分1〇2係在一約6〇的較低值， 然後在部分1〇4升到一值66，在一部分1〇6再回到—較 低值60。該較命SNR部分1〇4對應於讓人類眼睛最敏感 的頻率辄圍，因此，實際的實施型態可只在最須要的地方 提供額外準確度。 第2A圖繪不依據人類視覺系統（HVS)模型之調變臨 界點_)。此圖繪示出眼睛對—察視正弦波目標的最小檢 測能力。該Mt典型地係1 ^形且在典型的周遭亮度位 第6頁 5 10 請 先 m 讀 背 © 之 注 意 事 項 再 填 Jί裝 本 頁 ί I 訂 15 20 本紙張尺度綱巾S ©"^财（CNS)A4規格（21G x 297¾- 4 44 43 6 A7 B7 五、發明說明（4 ) 準下對介於3到5週期（cy)/等級（deg)的空間頻率最敏礅； 被定義為最大和最小明視度的差對最大和最小明视度 的和之比率。眼睛的對比靈敏度是該調變臨界點曲線的倒 數。標稱化之對比靈敏度是第2B圖所示眼睛的調變轉換 5函數（MTF);此一曲線顯示人類眼睛對於影像銳利度最敏 感時的空間影像頻率之範圍。 本發明使第2B圖的頻率範園匹配於第圖之SNR , 如此即產生一對於人類視覺最佳的ADC，同時使針對眼 睛較不敏感的頻率所需之解析度和準確度最小化。 10 第3圖係本發明的一個實施例之一方塊圖。一光源ι6〇 照射一目標景物162，反射光被一透鏡系統164導向一成 像器166，該成像器166提供一對於一水平線成光柵掃推 开> 式之串列波形；該訊號被一類比調整電路1〗2處理，其 典型地包含一相關雙取樣器（CDS);輸出波形係具有對應 15於該影像之一當時線條的頻率之一類比波形，然後此一輪 出被提供給一頻譜ADC114，其包含一 ADC116和一時間 平均插補電路Π8 ;頻譜ADCii4具有一如同第iB圖中 所說明者之SNR對空間頻率曲線。 頻譜ADC114之輸出被提供給數位後置處理電氣迴路 20 168，然後送到一影像顯示器170 ;該影像被一可能是人 類眼睛之一接收器172觀看，或者’ 一接收器可以是一例 如使用在一機器人系統中之光感測器。該接收器之Μτρ 對空間頻率響應’諸如在第2B圖中所示者，係匹配於頻 譜ADC114的SNR對空間頻率響應。 第7頁 本紙張尺度適用中國國家標準（CNS)A4規格（2】0 X 297公釐） Γ%先閱讀背面之迮意事項再填寫本頁) 裝----！|訂------!·'^--- 經濟部智慧財產局員工消費合作社印製 4 44 43 6 A7 B7 五 10 15 經濟部智慧財產局員工消費合作社印製 20 發明說明（5 ) 在一實施例中’本發明只針對在水平方向之空間頻 率，以保有成本效益。典型的影像感測器會水平地移出影 像資料以供處理。例如，大多數的影像和掃描系統感測器 (CCD、接觸影像感測器、MOS成像器）在一水平軸上移動 串列影像資料；然而，因為人類眼睛對於水平像差較敏感， 所以本發明便提供了足夠的改善。一個水平進行的間斷或 不連續（表示在垂直方向的一個錯誤1較能被大多數的人所 感受出’因為他們可感受具有由於光栅掃描而失真之週期 水平線條的商業電視之假頻混淆現象。 相同的技術可被應用在垂直方向以增加解析度：後來 的水平線條被以數位定義域儲存並產生额外的解析度。相 對於連續定址’這特別適合於可隨機定址之陣列。一個水 平連續位移的例子是傳統的CCD區域陣列成像器，其係 從爷一輸出端串列地移出資料，一次一條水平線β 一隨機 定址陣列的例子是一個MOS成像器，具有可以多種方式 被移出之可個別定址的像素。 插補電路例早 第4圖以箭頭1〇、a和14來繪示—ADC的三個不 同的連續輸出，此等樣本的次序是（η_丨）、η、（η+丨）；這 些樣本每一個的值分別為S(n-l)、S(n)和S(n+1卜在本發 明的一個實施例中，係假設實際的類比數值正線性地改 變，且因而晝出—條介於該（n-1)和該（n+1)樣本之值之間 的線16 ’該n樣本出現之交點被標示為a。 一臨界值18被設定，且該值S(n)被比較以看它是否 第8頁 (請先閱讀背面之注意事項再填寫本頁) 「紙張尺度適用侧家 444436 Α7 Β7 五、發明說明（6) 在該臨界值之範圍内。較佳地，唯若s(…是在該臨界值 範圍内，一額外之解析度位元才被設為i。 假如S(n)大於A，加入另一位元將只有使它更遠離預 期的直線，所以該額外的位元會被設為〇。假如s(n)小於 5臨界值18 ’並且臨界值18是1/2LSB，明顯的，實際樣 本正從該預期的相關值改變，所以該額外之解析度位元被 設為0。 第5圖係一繪示用以實行第4圖中所描述之計算的一 個方法之流程圖。首先，在步驟A中該A的值被決定為 10 S(n-I)和S(n+1)之平均；接著（步驟b)，S(n)的實際值被 和第4圖之臨界值18的較低部分比較，以判定它是否在 A以下大於1/2LSB ;假如它不是，該額外位元LSB則被 設為0(步驟C)。 假如S(n)大於該臨界值18的較低部分，就會被比較 15以看它是否小於A(步驟D);若否，那它即大於臨界值18 的較高部分，在步驟C中該LSB被設為〇，否則，該[SB 被設為1(步驟E) ’表示S(n)是在臨界值18的範圍内。 第6圖係用以實現本發明之邏輯電路的一個實施例之 方塊圖〇 —個1 〇位元ADC 20在線22接收一類比輸入並 20 在線24提供一個10位元輸出[S(n-l)];先前的輸出被儲 存在暫存器26和28’暫存器28儲存該s(n+l)值，而暫 存器26則儲存該S(n)值。 一加總電路30將該S(n+1)和S(n-l)值加總在一起。 因為兩個10位元數值被相加’結果可能是一具有進位之 第9頁 {請先閱讀背面之沒意事項再填寫本頁) 裝---- - - - - 訂-------炉 經濟部智慧財產局員工消費合作社印製 本紙張尺度適用中國國家標準（CNS)A4規格（2〗0 297公釐） 4 4 4 4 3 6 經濟部智慧財產局員工消費合作杜印製 Α7 Β7 五、發明說明（7 ) 11位元數值，較高的10位元，包括任何進位，在線34 上被提供給一比較器32，這些被和在線36上之s(n)位元 比較。藉由實行加總和將丨0個最高有效位元取出並將它 們往下移一位，在線34上之結果事實上即是s(n+1)和 5 s(n_1)的總和被除以2。假如該值在比較器32中未被判定 相等’這表示S(n)係不在1/2LSB的臨界值範圍内，而且 在線38上之一控制訊號會觸發開關4〇送出一零值以被提 供做為在線42上的額外解析度位元。 假如S(n)被判定為等於在線34上之結果，s(n)的值 10可能是在該臨界值之範圍内，且因此線38被作動以選擇 在線44上之該LSB »假如該LSB是1’這表示該除2平 均值是稍微高於S(n)，因此此一位元被送至線42以促 使該值S(n)更接近該平均值。假如在線44上之值LSB是 零’這表示對於解析度的這個位準而言，談平均值不大於 15 S(n)，所以S(n)不應該被增加；因此在線44上之零值被 放行直達線42做為額外位元β結果為在線46上解析度有 11個位元。 為了產生另一個位元的解析度，整個程序可以使用第 二組暫存器4 8、5 0和另一個加總電路5 2重覆執行》另外， 20 另一個比較器54會隨著另一個切換電路56而被使用到。 在某些應用中，當那個位元不是在臨界值範圍内時， 可能希望加上一雜訊成分到該解析度之額外位元。例如對 於一具有干擾的環境，這或許是有益的。它可能配上一些 校準程式的幫助’使它們更容易收斂。所以第6圖顯示_ 第10頁 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱） (請先閱績背面之注意事項再填寫本頁) 裝 !§J.— Ill — 11 — II — ^ ^ ------- i ^ ilnl —-- < Please read the notes on the back before filling out this page) 444 ^ 3 6 A7 B7 Employees ’Cooperatives, Intellectual Property Bureau, Ministry of Economic Affairs Printed 5. Description of the invention (3) I- 'Figure 2B is a graph showing a modulation transfer function versus space frequency. Figure 3 is a square circle of an exemplary embodiment of the present invention. Fig. 4 is a graph showing a linear interpolation method and a critical value according to an embodiment of the present invention. Fig. 5 is a flowchart illustrating one embodiment of the method of the present invention. Fig. 6 is a block diagram of one embodiment of a circuit for providing a resolution of two additional bits according to the present invention. Fig. 7 is a block diagram of the present invention for a single color or one mode of black-and-white light vision. Fig. 8 is a block diagram of the second mode of the present invention for multiple colors or different spatial frequencies. Figure 1A shows the value of the signal-to-noise ratio (stupid) vs. frequency of the analog-to-digital converter as a curve i 〇. It can be seen from the figure that the European spoon is maintained at 66 until a very high frequency, that is, essentially constant in this range. The first drawing is a drawing of FIG. 1A used in the present invention. As can be seen, the SNR initially starts at a low value of about 60 in part 102, then rises to a value of 66 in part 104, and then returns to a lower value 60 in part 106. The fatal SNR portion 104 corresponds to the frequency range that makes the human eye most sensitive, so the actual implementation can provide additional accuracy only where it is most needed. Figure 2A illustrates the modulation threshold of the human visual system (HVS) model. This figure illustrates the minimum detection capability of eye-to-sight sine wave targets. The Mt is typically 1 ^ shaped and in the typical surrounding brightness level. Page 6 5 10 Please read the precautions of © © before filling in this page. I order 15 20 paper standard outline S © " ^ (CNS) A4 specification (21G x 297¾- 4 44 43 6 A7 B7 V. Description of the invention (4) The most sensitive to the spatial frequency between 3 to 5 cycles (cy) / grade (deg); is defined; Is the ratio of the difference between the maximum and minimum vision to the sum of the maximum and minimum vision. The contrast sensitivity of the eye is the inverse of the modulation threshold curve. The nominal contrast sensitivity is the modulation of the eye shown in Figure 2B. 5 function (MTF); This curve shows the range of the spatial image frequency when the human eye is most sensitive to the sharpness of the image. The present invention matches the frequency range of Figure 2B to the SNR of Figure 2. The best ADC for human vision, while minimizing the resolution and accuracy required for frequencies less sensitive to the eye. 10 Figure 3 is a block diagram of one embodiment of the present invention. A light source illuminates a target Scene 162, the reflected light is directed to 10% by a lens system 164 166, the imager 166 provides a tandem waveform of a raster sweep for a horizontal line >type; the signal is processed by an analog adjustment circuit 1〗 2 which typically includes a correlated double sampler (CDS); The output waveform has an analog waveform corresponding to a frequency of 15 lines at that time in the image, and then this round is provided to a spectrum ADC114, which includes an ADC116 and a time-average interpolation circuit Π8; the spectrum ADCii4 has an image like The SNR vs. spatial frequency curve illustrated in the iB diagram. The output of the spectrum ADC114 is provided to a digital post-processing electrical circuit 20 168, and then sent to an image display 170; the image is received by a receiver 172, which may be one of the human eyes Watch, or 'a receiver may be a light sensor, such as used in a robotic system. The receiver's Mτρ response to spatial frequency', such as that shown in Figure 2B, matches the SNR of the spectrum ADC114 Response to space frequency. Page 7 This paper size applies Chinese National Standard (CNS) A4 specification (2) 0 X 297 mm. Γ% Please read the intention on the back before filling in this page) Install ----! Order ------! · '^ --- Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 4 44 43 6 A7 B7 5 10 15 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 20 Invention description (5) In one embodiment, 'the present invention only targets spatial frequencies in the horizontal direction to maintain cost effectiveness. A typical image sensor moves the image data horizontally for processing. For example, most image and scanning system sensors (CCD, contact image sensor, MOS imager) move serial image data on a horizontal axis; however, because human eyes are more sensitive to horizontal aberrations, The invention provides sufficient improvement. A horizontal discontinuity or discontinuity (indicating an error in the vertical direction1 is more likely to be felt by most people 'because they can feel the aliasing of commercial televisions with periodic horizontal lines distorted by raster scanning The same technique can be applied in the vertical direction to increase resolution: Later horizontal lines are stored in digital domains and generate additional resolution. This is particularly suitable for random addressable arrays compared to continuous addressing. A horizontally continuous An example of displacement is a traditional CCD area array imager, which removes data serially from the output of the Yeyi, one horizontal line β at a time. An example of a random addressing array is a MOS imager, which can be removed in various ways. The address of the pixel. An example of the interpolation circuit is shown in Figure 4 with arrows 10, a, and 14—the three different continuous outputs of the ADC. The order of these samples is (η_ 丨), η, (η +丨); the values of each of these samples are S (nl), S (n), and S (n + 1) In an embodiment of the present invention, the actual analog values are assumed Change linearly, and thus day out—a line between the values of the (n-1) and the (n + 1) samples 16 'The intersection of the n samples appears labeled a. A critical value 18 is Set, and the value S (n) is compared to see if it is on page 8 (please read the precautions on the back before filling this page) "Paper size applicable side home 444436 Α7 Β7 V. Description of invention (6) At this threshold Value range. Preferably, an additional resolution bit is set to i only if s (... is within the critical value range. If S (n) is greater than A, adding another bit will only have Make it further away from the expected straight line, so the extra bit will be set to 0. If s (n) is less than 5 critical value 18 'and critical value 18 is 1 / 2LSB, obviously, the actual sample is moving from the expected The correlation value changes, so the additional resolution bit is set to 0. Figure 5 is a flowchart showing a method for performing the calculations described in Figure 4. First, in step A, the A The value of S (nI) is determined as the average of 10 S (nI) and S (n + 1); then (step b), the actual value of S (n) is compared with the lower part of the critical value 18 in Fig. 4 To determine whether it is greater than 1/2 LSB below A; if it is not, the extra bit LSB is set to 0 (step C). If S (n) is greater than the lower part of the threshold 18, it will 15 is compared to see if it is less than A (step D); if not, then it is greater than the higher part of the critical value 18, the LSB is set to 0 in step C, otherwise, the [SB is set to 1 ( Step E) 'Indicates that S (n) is in the range of critical value 18. Figure 6 is a block diagram of an embodiment of a logic circuit used to implement the present invention. A 10-bit ADC 20 receives a line 22 Analog input and 20 provide a 10-bit output [S (nl)] on line 24; the previous output is stored in registers 26 and 28. Register 28 stores the value of s (n + l), while register 26 stores the S (n) value. A summing circuit 30 adds up the S (n + 1) and S (n-1) values together. Because the two 10-bit values are added, 'the result may be a page 9 with a carry (please read the unintentional matter on the back before filling out this page) Loading --------Order ----- -Printed by the Intellectual Property Bureau's Consumer Cooperatives of the Ministry of Economics This paper is printed in accordance with the Chinese National Standard (CNS) A4 specification (2〗 0 297 mm) 4 4 4 4 3 6 Α7 Β7 V. Description of the invention (7) 11-bit value, the higher 10-bit value, including any carry, is provided to a comparator 32 on line 34, and these are compared with the s (n) bit on line 36 . By implementing the sum and taking out the 0 most significant bits and moving them down one bit, the result on line 34 is actually the sum of s (n + 1) and 5 s (n_1) divided by 2 . If the value is not determined to be equal in the comparator 32, this means that S (n) is not within the threshold of 1 / 2LSB, and one of the control signals on line 38 will trigger the switch 4 to send a zero value to be provided. As additional resolution bits on line 42. If S (n) is determined to be equal to the result on line 34, the value of s (n) 10 may be within the threshold, and therefore line 38 is actuated to select the LSB on line 44 »if the LSB Yes 1 'This means that the division by 2 average is slightly higher than S (n), so this bit is sent to line 42 to push the value S (n) closer to the average. If the value LSB on line 44 is zero, this means that for this level of resolution, the average value is not greater than 15 S (n), so S (n) should not be increased; therefore, the zero value on line 44 The cleared direct line 42 as the extra bit β results in 11 bits of resolution on the line 46. In order to generate another bit resolution, the entire program can be executed repeatedly using a second set of registers 48, 50, and another summing circuit 5 2. In addition, 20 another comparator 54 will follow the other A switching circuit 56 is used. In some applications, when that bit is not within a critical range, it may be desirable to add a noise component to the extra bits of the resolution. For example, this may be beneficial for a noisy environment. It may be accompanied by some calibration programs ’to make them easier to converge. Therefore, Fig. 6 shows _ page 10 This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 public love) (please read the precautions on the back of the results before filling this page)
V 44443 6 Α7 Β7 五、發明說明（8) 由來自比較器54之依意願附加的控制線6〇控制之可選配 顫動電路58 ^當控制線60指出零值應被選擇時，換成該 電路5 8被選擇到，以置放一假隨機數值（丨或〇)在線62 上以給予線64上之輸出第12個位元。假如需要，一顏動 5 電路59也可以被用來得到該解析度之第u個位元。 因此’本發明之此一實施例提供一不需要相同解析度 的一個標準ADC的高精密度元件之改良解析度aDC。已 確定最大訊號降級是2Λ當中的一個碼，本發明可以用比一 使用類比元件以直接從類比訊號產生額外解析度的額外位 10 元之等效實施型態所使用者更少的矽面積來實現，本發明 改善ADC對於低頻訊號的訊號雜訊比，而不會使高頻訊 號大幅惡化。如上所述，這些優點係針對例如可能在影像 資料訊號中遇到者的局部相關訊號而得到β 第7圖是一本發明之一單一色彩模式的實施型態示意 15 圖。類比調整電路120提供一訊號至一個ADC 122，然後 一訊號被提供給一可被插補電路126存取之儲存電路 124。 第8圖顯示一針對三種不同顏色之替換模式。該三種 顏色可在各別類比調整電路128、130和132輸入，然後 20 該等訊號經由可規劃增益放大器（PGA) 134、136和138被 提供給一多工器14〇。此等顏色訊號經由同一 ADCM2被 輪流地提供並分散到個別的儲存電路144、146和148， 它們將在此被插補電路150、152和1 54存取以進行插補。 替換地，其他的實施型態是可能的。例如，一拜耳圖 第11頁 本紙張尺度過用令國國家標準（CNS)A4規格（210 X 297公楚） (請先閱讀背面之注意事項再填寫本頁) 裝--------訂---------产 經濟部智慧財產局員工消費合作社印製 444 43 6 A7 B7 經 濟 部 智 慧 財 產 局 員 工 消 費 合 h 社 印 % 五、發明說明（9 ) 案可被提供給取樣CDS電路或一取樣保持電路，然後經 由一 PGA和多工器提供至一單一 ADCe替換地G(綠）和 藍/紅顏色可被同時取樣並提供給一取樣保持（電路）或 CDS，然後接著經由一 PGA提供給一多工器並進入一單 5 > ADC。另一例子可提供一經由一 PGA和多工器被提供 進入一單一 ADC之三色串列輸出取樣保持或CDS。其他 實施型態對於熟知本項技藝之人士將是顯而易知的（於此 不加贅述）。 本發明也可被提供以藉由從垂直鄰接影像資料插補額 外的準確度而產生增加的垂直解析度和準確度，然而，此 一實施例須要更多電路以儲存影像資料。 在實施例中，較尚解析度被提供之諸頻率的範圍可 就讀每一種顏色加以改變。例如眼睛係典型地對於藍色的 變化較不敏感，因此藍色頻率的一個較狹窄的範圍可被提 供較南的解析度。替換地，藍色的所有色彩可用一低解析 度提供，而例如綠色和紅色則是完全地或大部分地以一較 南解析度提供。 用以提供本發明之頻譜響應的一些替換實施例可被使 用。例如一帶通濾波器可被放在一高解析度ADC：的前面。 替換地，一帶通數位濾波器可放在一個ADC後使用；例 如一 10位元寬之訊號路徑可被使用在低解析度，切換到 一供高解析度用之12位元寬路徑。 如同將被熟知本項技藝之人士所了解者，本發明可用 /、他特殊开;ί式實施且不會脫離它的基本特性。例如當一額 L____ 第12頁 本紙張尺度適財關家標準（c㈣----- 10 請 先 閱 讀 背 © 之 注 意 事 項 再 填 寫 本 頁 裝 I I I I I 訂 15 20V 44443 6 Α7 Β7 V. Description of the invention (8) The optional dithering circuit 58 controlled by the control line 60 added from the comparator 54 as desired. ^ When the control line 60 indicates that the zero value should be selected, replace it with this. Circuit 58 is selected to place a pseudo-random value (丨 or 〇) on line 62 to give the twelfth bit of output on line 64. If necessary, the Y5 circuit 59 can also be used to obtain the u-th bit of the resolution. Therefore, this embodiment of the present invention provides an improved resolution aDC of a high precision component of a standard ADC that does not require the same resolution. It has been determined that the maximum signal degradation is a code of 2Λ. The present invention can use less silicon area than an equivalent implementation of 10 yuan using an analog component to directly generate additional bits of additional resolution from the analog signal. Realized, the present invention improves the signal-to-noise ratio of the ADC for low-frequency signals without significantly deteriorating the high-frequency signals. As described above, these advantages are obtained by, for example, local correlation signals of a person who may be encountered in the image data signal. FIG. 7 is a schematic diagram of an embodiment of a single color mode of the present invention. The analog adjustment circuit 120 provides a signal to an ADC 122, and then a signal is provided to a storage circuit 124 which can be accessed by the interpolation circuit 126. Figure 8 shows a replacement pattern for three different colors. The three colors can be input in the respective analog adjustment circuits 128, 130, and 132, and then these signals are supplied to a multiplexer 14 through a programmable gain amplifier (PGA) 134, 136, and 138. These color signals are alternately provided through the same ADCM2 and distributed to the individual storage circuits 144, 146, and 148, and they will be accessed here by the interpolation circuits 150, 152, and 154 for interpolation. Alternatively, other implementation forms are possible. For example, page 11 of the Bayer chart. This paper is used in the national standard (CNS) A4 specification (210 X 297 cm). (Please read the precautions on the back before filling this page.) -Order --------- Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Industry and Economics 444 43 6 A7 B7 Employee Consumption Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs% Printed by the Ministry of Economic Affairs (9) The case can be provided Giving a sample CDS circuit or a sample-and-hold circuit and then supplying it to a single ADCe via a PGA and multiplexer G (green) and blue / red colors can be sampled simultaneously and provided to a sample-and-hold (circuit) or CDS, It is then supplied to a multiplexer via a PGA and enters a single 5 > ADC. Another example could provide a three-color serial output sample-and-hold or CDS that is provided via a PGA and multiplexer into a single ADC. Other implementation types will be obvious to those who are familiar with this technique (no further details here). The present invention can also be provided to generate increased vertical resolution and accuracy by interpolating additional accuracy from vertically adjacent image data, however, this embodiment requires more circuitry to store the image data. In an embodiment, the range of frequencies to which the higher resolution is provided may be changed for each color. For example, the eye system is typically less sensitive to changes in blue, so a narrower range of blue frequencies can provide a more southern resolution. Alternatively, all colors of blue can be provided at a low resolution, while for example green and red are provided completely or mostly at a souther resolution. Alternative embodiments to provide the spectral response of the present invention can be used. For example, a band-pass filter can be placed in front of a high-resolution ADC :. Alternatively, a band-pass digital filter can be used after an ADC; for example, a 10-bit wide signal path can be used at a low resolution and switched to a 12-bit wide path for high resolution. As will be understood by those skilled in the art, the present invention can be implemented specifically and without departing from its basic characteristics. For example, when an amount of L____ page 12 This paper is suitable for financial standards (c㈣ ----- 10)
10 範圍 -4 44 43 6 發明說明（10 外位元的解析度針對一個s(n)訊號而被決定時，當那個 同樣的訊號為後來的s(n+1)訊號時，則那個额外位元可 被使用，並可被用來計算平均值。這可被用來當做一在同 一時間產生兩個額外位元的解析度之替代方法，而不必重 覆用以得到第-個位元之解析度的步驟。在其他實施例 中’曲線而非-直線，可被使用，且使用超過兩個樣本以 趨近此一曲線的一個程序可被包含在計算過程中。可替換 地，其它臨界值可被使用。所以，前面的描述之用意在於 說明而不是要限制下列申請專利範圍中所描述之本發明的 (請先閱讀背面之注意事項再填寫本頁) 5 經濟部智慧財產局員工消費合作社印製 ο 2 儿件標號對照表： 100 曲線 102 部分 104 部分 106 部分 160 光源 162 目標景物 164 透鏡系統 166 成像器 Π2 類比調整電路 114 頻譜ADC Π6 ADC 118 時間平均插補電路 168 數位後置處理電氣迴路 170 影像顯示器 172 接收器 10 箭頭 12 箭頭 14 箭頭 16 線 18 臨界值 22 線 24 線 26 暫存器 28 暫存器. 30 加總電路 34 線 第13頁 本紙張又度適用中國國家標準（CNS)A4規格（210 X 297公S ) ----訂---------'^ν· 4 4 d 4 3 8 經濟部智慧財產局員工消費合作社印製 A7 B7 五、發明說明（11 ) 32 比較器 36 線 40 開關 42 線 38 線 44 線 52 加總電路 54 比較器 5 '56 切換電路 60 控制線 58 可選配顫動電路 62 線 64 線 59 顫動電路 120 類比調整電路 122 ADC 126 插補電路 124 儲存電路 10 128 類比調整電路 130 類比調整電路 132 類比調整電路 134 可規劃增益放大器 136 可規劃增益放大器 138 可規劃增益放大器 140 多工器 142 ADC 150 插補電路 152 插補電路 15 154 插補電路 144 儲存電路 146 儲存電路 148 儲存電路 第14頁 1--------I! 裝-------1 訂-------- ·%Ί (請先閱讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐）10 range-4 44 43 6 Description of the invention (When the resolution of the 10 outer bits is determined for an s (n) signal, when the same signal is a subsequent s (n + 1) signal, the extra bit The element can be used and used to calculate the average. This can be used as an alternative to generating a resolution of two extra bits at the same time, without having to repeat it to get the first bit Resolution step. In other embodiments, 'curve instead of-straight line' can be used, and a program using more than two samples to approach this curve can be included in the calculation process. Alternatively, other critical Values can be used. Therefore, the foregoing description is intended to illustrate rather than limit the invention described in the scope of patent applications below (please read the notes on the back before filling this page) Printed by the cooperative ο 2 child parts reference table: 100 curve 102 part 104 part 106 part 160 light source 162 target scene 164 lens system 166 imager Π 2 analog adjustment circuit 114 spectrum ADC Π 6 ADC 118 Intermediate average interpolation circuit 168 Digital post-processing electrical circuit 170 Video display 172 Receiver 10 Arrow 12 Arrow 14 Arrow 16 Line 18 Threshold 22 Line 24 Line 26 Register 28 Register 30 30 Sum Circuit 34 Line 13 This paper is again applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 male S) ---- Order --------- '^ ν · 4 4 d 4 3 8 Intellectual Property Bureau of the Ministry of Economic Affairs Printed by employee consumer cooperative A7 B7 V. Description of invention (11) 32 Comparator 36 line 40 switch 42 line 38 line 44 line 52 total circuit 54 comparator 5 '56 switching circuit 60 control line 58 optional vibration circuit 62 line 64 line 59 dithering circuit 120 analog adjustment circuit 122 ADC 126 interpolation circuit 124 storage circuit 10 128 analog adjustment circuit 130 analog adjustment circuit 132 analog adjustment circuit 134 programmable gain amplifier 136 programmable gain amplifier 138 programmable gain amplifier 140 multiplexer 142 ADC 150 Interpolation circuit 152 Interpolation circuit 15 154 Interpolation circuit 144 Storage circuit 146 Storage circuit 148 Storage circuit Page 14 1 -------- I! Equipment --------- 1 Order-- - ----- ·% Ί (Please read the notes on the back before filling out this page) This paper size is applicable to China National Standard (CNS) A4 (210 X 297 mm)