1334304 099年07月23日修正替換頁 六、發明說明： 【發明所屬之技術領域】 [0001] 本發明涉及攝像技術，特別涉及一種圖像處理方法及圖 像處理裝置。 ’ 【先前技術】 [0002] 請參閱圖6，為攝像裝置之影像感測器產生之圖像之一種 灰階（gray scale)分佈圖（histogram)。其中，坐標軸 之橫軸表示灰階，0表示純黑，^表示純白，靠近0處， 灰階低，為圖像暗部，靠近D處，灰階高，為圖像明部； 縱軸表示處於對應灰階之像素個數；曲線99為圖像之灰 階分佈曲線。曲線99描述：在以下可能之情況：1)場景 過暗；2)曝光不足；3)場景明暗反差過大，為防止描 述場景明部之像素發生高光溢出（blooming)而出現浸潤 (smear)現象，採用小曝光值進行曝光。攝像裝置攝得圖 像過暗，.圖像暗部細節缺失嚴重。因此，需對圖像進行 處理，拉升圖像暗部像素之灰階，以得到明暗適合、層 次豐富之圖像。 [0003] 為此，傳統之圖像處理方法在處理圖像時，放大圖像暗 部像素之訊號量（電量），以此拉升圖像訊號暗部像素灰 階。然而，像素之訊號量總包括各種各樣之雜訊（如白雜 訊或暗電流）。傳統圖像處理方法在放大圖像訊號暗部像 素灰階之同時，訊號量中之雜訊對應放大，導致圖像品 質嚴重劣化。 【發明内容】 [0004] 有鑒於此，有必要提供一種可避免圖像品質劣化之圖像 096130562 表單編號A0101 第4頁/共20頁 0993265692-0 13343041334304 Revised replacement page on July 23, 099. Description of the Invention: [Technical Field] [0001] The present invention relates to an image pickup technique, and more particularly to an image processing method and an image processing apparatus. [Prior Art] [0002] Please refer to FIG. 6, which is a gray scale histogram of an image generated by an image sensor of a camera. Wherein, the horizontal axis of the coordinate axis represents gray scale, 0 represents pure black, ^ represents pure white, close to 0, gray scale is low, is dark part of image, close to D, gray level is high, is the bright part of the image; vertical axis indicates The number of pixels in the corresponding gray level; curve 99 is the gray scale distribution curve of the image. Curve 99 describes: in the following possible situations: 1) the scene is too dark; 2) the exposure is insufficient; 3) the contrast between the scene and the light is too large, in order to prevent the smear phenomenon from appearing in the highlighting of the pixels in the bright part of the scene, Exposure is performed with a small exposure value. The image taken by the camera is too dark, and the details of the dark part of the image are seriously missing. Therefore, the image needs to be processed to increase the gray level of the dark portion of the image to obtain an image with a suitable brightness and darkness. [0003] For this reason, the conventional image processing method enlarges the signal amount (electric quantity) of the dark portion of the image when processing the image, thereby pulling up the gray level of the dark portion of the image signal. However, the amount of signal in a pixel always includes a variety of noise (such as white noise or dark current). The conventional image processing method amplifies the gray level of the dark portion of the image signal, and the noise in the signal amount is enlarged correspondingly, resulting in severe deterioration of the image quality. SUMMARY OF THE INVENTION [0004] In view of this, it is necessary to provide an image that can avoid image quality degradation 096130562 Form No. A0101 Page 4 of 20 0993265692-0 1334304
099年07月23日修正替换W 處理方法及圖像處理裝置。 [0005] 一種用於攝像裝置之圖像處理方法，該方法包括如下步 驟： [0006] 轉換該攝像裝置產生之圖像，生成包括亮度分量之原圖 像； [0007] 複製該原圖像之亮度分量，生成副圖像； [0008] 反轉該副圖像，生成負圖像； [0009] 去除該負圖像之雜訊，生成去噪負圖像；及 [0010] 將預定比例值之去噪負圖像合.併入該原圖像之亮度分量 ，生成目標圖像。 [0011] 一種用於攝像裝置之圖像處理裝置，其包括： [0012] 轉換模組，用於轉換該攝像裝置產生之圖像，生成包括 亮度分量之原圖像； [0013] 複製模組，用於複製該原圖像之亮度分量，生成副圖像 9 [0014] 反轉模組，用於反轉該副圖像，生成負圖像； [0015] 去嗓模組，用於去除該負圖像之雜訊，生成去噪負圖像 ;及 [0016] 合併模組，用於將預定比例值之去噪負圖像合併入原圖 像之亮度分量，生成目標圖像。 [0017] 相較於習知技術，該圖像處理方法及該圖像處理裝置通 096130562 表單編號A0101 第5頁/共20頁 0993265692-0 1334304 099年 07月 23 日^ 過對用於拉升圖像暗部像素灰階值之負片進行去嗓處理 ，避免在拉升圖像暗部像素灰階值之同時，放大圖像雜 訊，劣化圖像。 【實施方式】 [0018] 請參閱圖1，較佳實施例之圖像處理裝置1〇〇設置於攝像 裝置（圖未示）之影像感測器200後端，用於處理影像感測 器200產生之圖像。影像感測器200可以係電荷耦合器 (Charge Coupled Device, CCD)或補充性半導體 (Complementary Metal Oxide Semiconductor, CMOS)裝置，作為範例，本實施例之影像感測器200為 CCD。 [0019] 圖像處理裝置100包括轉換模組10、複製模組20、反轉模 組30、去噪模組40及合併模組50。轉換模組1〇用於轉換 影像感測器200產生之圖像，生成包括亮度分量之原圖像 。複製模組20用於複製原圖像之亮度分量，生成副圖像 。反轉模組30用於反轉副圖像，生成負圖像。去噪模組 30用於去除負圖像之雜訊，生成去噪負圖像。合併模組 50用於將預定比例值之去噪負圖像合併入原圖像之亮度 分量，生成目標圖像。 [0020] [0021] 具體地’圖像處理裝置1〇〇還包括存儲模組60，用於保存 原圖像’合併模組50從存儲模組60獲取原圖像。 請參閱圖2 ’較佳實施例之圖像處理方法包括如下步驟： S01 .轉換電荷耦合器2〇〇產生之圖像，生成包括亮度分 量之原圖像； 096130562 表單編號A0101 第6頁/共20頁 0993265692-0 [0022] 1334304 099年07月23日修正替換頁 [0023] S02 :複製原圖像之亮度分量，生成副圖像； [0024] S03 :反轉副圖像，生成負圖像； [0025] S04:去除負圖像之雜訊，生成去噪負圖像；及 . [0026] S05 :將預定比例之去噪負圖像合併入原圖像之亮度分量 ，生成目標圖像。 [0027] 以下結合圖3，進一步說明較佳實施例之圖像處理方法。 [0028] 圖3中，R、Y、Yc、I、A及F分別為影像感測器200產生 之圖像、原圖像、副圖像、負圖像、去噪負圖像及目標 圖像。 [0029] 通常，影像感測器200採用三與感光單元（S未示）感測圖 像R第m列，第η行（m，η為自然數，下同>i像素之紅色 分量R 、綠色分量G 及藍色分量B ，並採用紅色分量 mn mn mn Rmn、綠色分量G 及藍色分量B 表徵圖像R第m列，第η mn mn 行之像素。即，影像感測器200以熟知之全彩模式輸出圖 像R。Y 及F 分別為原圖像Y及目標圖像？第111列，第η行 mn mn 之像素之亮度分量（灰階值表'示）°Yc ' I 及A 分別 mn mn mn 為副圖像Yc、負圖像I及去噪負圖像A第in列，第η行之像 素之灰階值。作為範例，本實施例之圖像R、原圖像Υ、 副圖像Yc、負圖像I、去噪負圖像Α及目標圖像F包括3x3 個像素。 [0030] 具體地，對於步驟S01(轉換步驟），轉換模組10採用公式 71 ： [0031] Y =0.299R +0.587G +0.114B mn mn mn mn 096130562 表單編號A0101 第7頁/共20頁 0993265692-0 1334304 099年07月23日梭正替換百 [0032] [0033] [0034] [0035] [0036] [0037] [0038] [0039] [0040] [0041] [0042] [0043] [0044] 096130562 轉換圖像R，生成原圖像Y。當然，步驟S01並不限於本實 施例，轉換模組10可採用其他演算法轉換圖像。 更加具體地，該方法在步驟S01(轉換步驟）後還包括步驟 S12 :保存原圖像Y。 將原圖像Y保存於存儲模組60，以供後續步驟讀取、使用 0 對於步驟S02(複製步驟），複製模組20採用公式72 :Correction of the replacement W processing method and image processing apparatus on July 23, 099. [0005] An image processing method for an image pickup apparatus, the method comprising the following steps: [0006] converting an image generated by the image pickup device to generate an original image including a brightness component; [0007] copying the original image a luminance component, generating a secondary image; [0008] inverting the secondary image to generate a negative image; [0009] removing noise of the negative image to generate a denoised negative image; and [0010] setting a predetermined ratio The denoised negative image is combined with the luminance component of the original image to generate a target image. [0012] An image processing apparatus for an image pickup apparatus, comprising: [0012] a conversion module, configured to convert an image generated by the image pickup apparatus, and generate an original image including a brightness component; [0013] a copy module For copying the luminance component of the original image, generating a secondary image 9 [0014] an inversion module for inverting the secondary image to generate a negative image; [0015] removing the module for removal The noise of the negative image generates a denoised negative image; and [0016] a merge module is configured to merge the denoised negative image of the predetermined ratio into the luminance component of the original image to generate a target image. [0017] Compared with the prior art, the image processing method and the image processing apparatus pass 096130562 Form No. A0101 Page 5 / Total 20 Page 0993265692-0 1334304 099 July 23 ^ Over the pair for pulling up The negative of the grayscale value of the dark portion of the image is processed to avoid magnifying the image noise and degrading the image while lifting the grayscale value of the dark portion of the image. [Embodiment] Referring to FIG. 1, an image processing apparatus 1 of a preferred embodiment is disposed at a rear end of an image sensor 200 of an imaging device (not shown) for processing the image sensor 200. The resulting image. The image sensor 200 can be a Charge Coupled Device (CCD) or a Complementary Metal Oxide Semiconductor (CMOS) device. For example, the image sensor 200 of the present embodiment is a CCD. [0019] The image processing apparatus 100 includes a conversion module 10, a copy module 20, a reverse module 30, a noise canceling module 40, and a merge module 50. The conversion module 1 is configured to convert an image generated by the image sensor 200 to generate an original image including a luminance component. The copy module 20 is for copying the luminance component of the original image to generate a sub-image. The inversion module 30 is for inverting the sub-image to generate a negative image. The denoising module 30 is used to remove the noise of the negative image and generate a denoised negative image. The merge module 50 is configured to merge the denoised negative image of the predetermined scale value into the luminance component of the original image to generate a target image. [0021] Specifically, the image processing apparatus 1 further includes a storage module 60 for storing the original image. The merge module 50 acquires the original image from the storage module 60. Referring to FIG. 2, the image processing method of the preferred embodiment includes the following steps: S01. Converting the image generated by the charge coupler 2 to generate an original image including a luminance component; 096130562 Form No. A0101 Page 6 / Total 20 pages 0993265692-0 [0022] 1334304 Correction replacement page of July 23, 099 [0023] S02: copying the luminance component of the original image to generate a secondary image; [0024] S03: inverting the secondary image to generate a negative image [0025] S04: removing the noise of the negative image to generate a denoised negative image; and [0026] S05: incorporating a predetermined proportion of the denoised negative image into the luminance component of the original image to generate a target image image. [0027] The image processing method of the preferred embodiment will be further described below with reference to FIG. 3. [0028] In FIG. 3, R, Y, Yc, I, A, and F are respectively an image generated by the image sensor 200, an original image, a sub image, a negative image, a denoised negative image, and a target image. image. [0029] Generally, the image sensor 200 senses the mth column of the image R by using three photosensitive cells (not shown), and the nth row (m, η is a natural number, the same as the red component R of the i pixel) , the green component G and the blue component B, and the red component mn mn mn Rmn, the green component G and the blue component B are used to represent the mth column of the image R, the pixel of the η mn mn row. That is, the image sensor 200 The image R is output in the well-known full color mode. Y and F are the original image Y and the target image respectively. The 111th column, the luminance component of the pixel of the ηth row mn mn (the gray scale value table 'shows) °Yc ' I and A respectively mn mn mn are the sub-picture Yc, the negative image I, and the gray-scale value of the pixel in the nth column and the n-th row of the denoising negative image A. As an example, the image R of the present embodiment, the original The image Υ, the sub image Yc, the negative image I, the denoising negative image Α, and the target image F include 3×3 pixels. [0030] Specifically, for step S01 (conversion step), the conversion module 10 adopts a formula 71 : [0031] Y = 0.299R +0.587G +0.114B mn mn mn mn 096130562 Form No. A0101 Page 7 / Total 20 Page 0993265692-0 1334304 On July 23, 2009, Shuttle is replacing 100 [0032] [003 [0034] [0037] [0038] [0038] [0040] [0044] [0044] 096130562 Convert image R, generate original image Y. Of course, steps S01 is not limited to the embodiment, and the conversion module 10 may use other algorithms to convert the image. More specifically, the method further includes a step S12 after the step S01 (conversion step): saving the original image Y. Y is stored in the storage module 60 for subsequent steps to read, using 0 for step S02 (copying step), and the copy module 20 uses equation 72:
Yc =Y mn mn 複製原圖像Y之亮度分量，生成副圖像Yc。 對於步驟S03(反轉步驟），反轉模組30採用公式73 : I =D-Yc mn mn 反轉副圖像Yc，生成負圖像I。其中，D為自然數，表示 副圖像Yc之最高灰階值，如常見之256階或4096階（即 D=255或D=4095)。 對於步驟S04(去噪步驟），去噪模組40採用公式74 : A =S04(I ) mn mn 去除負圖像I之雜訊，生成去噪負圖像A。其中， S04(I )表示去脅模組40對像素I 之處理，如，去嚼模 mn mn 組40採用均值濾波器或中值濾波器，對負圖像I進行平均 濾波（法）或中值濾波（法）。平均濾波（法）或中值濾波（法 )演算法簡單，可縮短圖像處理時間。 表單編號A0101 第8頁/共20頁 0993265692-0 1334304 [0045] [0046] [0047] [0048] [0049] [0050] [0051] [0052] 可替換地，可採用改進型均值濾波器或中值濾波器執行 步驟S04(去噪步驟）以改善負圖像丨之去噪效果。 凊參閱圖4，優選地，本實施例之步驟s〇4(去噪步驟）包 括如下子步驟： 5041 :劃分負圖像I為複數區域； 5042 .计昇每個區域之灰階平均值； 5043 .判斷每個像素之灰階值是否超出所在區域之灰階 平均值之預定範圍；及 5044 .替換灰階值超出對應灰階平均值預定範圍之像素 之灰階值為對應灰階平均值纟 & 如此，通過子步驟S〇43(判斷步驟）判斷\^素是否為噪化 像素（即灰階值超出所在區域之灰階平均值預定範園之像 素），並僅對噪化像素作平均濾波。既可濾掉雜訊，又可 保留正$像素之灰階特徵值，避免去噪後負圖像I變模糊 其中，預定範圍可依據使用者之喜好或者經驗設置。更 優選地，預定範圍可由去噪模組4〇判斷決定。參考方案 為：子步驟S041(計算步驟）還計算每個區域之灰階方差 ，一個像素對應之預定範圍取決於所在區域之灰階方差 ，如，二倍方差或三倍方差。如此，依據概率統計原理 :絕大部分資訊分佈於資訊均值上下二倍方差或三倍方 差内，可將灰階值超出所在區域之灰階平均值預定範圍（ 二倍方差或三倍方差）之像素視為噪化像素。 096130562 表單編號A0101 第9頁/共20頁 0993265692-0 1334304 099年07月23日修正替換頁 [0053] 對於步驟S05(合併步驟），合併模組50採用公式75 : [0054] F =S05(A ) mn mn [0055] 將預定比例之去噪負圖像A合併入原圖像Y之亮度分量， 生成目標圖像F。其中，S05(A )表示合併模組50對像素 mn A 之處理。具體地，為保留原圖像Y之亮部特性，合併 mn 模組50應避免對原圖像Y亮部像素之合併。故， S05(A )(合併步驟）包括如下子步驟： mn [0056] S051 :判斷原圖像Y之每個像素之灰階值是否小於預設域 值，並約定灰階值小於預設域值之像素為暗部像素； [0057] S052 :計算去噪負圖像A中對應原圖像Y各個暗部像素之 各個像素之灰階值之預定比例值；及 [0058] S053 :將各個預定比例值之灰階值加入原圖像Y對應暗部 像素之灰階值。 [0059] 預設域值可預先依據使用者之喜好或經驗設定。可以理 解，原圖像Y暗部像素數目越少，合併模組50之計算量越 小，圖像處理時間縮短。故，優選地，預設域值之設定 應使原圖像Y暗部像素數目小於預定數目。 [0060] 預定比例值可預先依據使用者之喜好或經驗設定。具體 地，該方法在步驟S05(合併步驟）前還包括步驟： [0061] S0 0 :設定預定比例值。 [0062] 較理想之預定比例值可通過實驗方法得出，並内建於攝 像裝置内，以供使用者選定。本實施之預定比例值包括 10%、15%、18%、20%及25%。 096130562 表單編號A0101 第10頁/共20頁 0993265692-0 1334304 099年07月23日梭正替換頁 [0063] 由於合併模組50並非對原圖像Y之所有像素進行合併（僅 合併暗部像素）。為降低去噪模組40之計算量，縮短圖像 處理時間。優選地，步驟S043(判斷步驟）可僅判斷負圖 像I中對應原圖像Y暗部像素之之各個像素之灰階值是否 超出所在區域之灰階平均值之預定範圍。 [0064] 綜上所述，本發明確已符合發明專利要件，爰依法提出 專利申請。惟，以上所述者僅為本發明之較佳實施方式 ，舉凡熟悉本案技藝之人士，於援依本案發明精神所作 之等效修飾或變化，皆應包含於以下之申請專利範圍内 〇 【圖式簡單說明】 ： [0065] 圖1為較佳實施例之圖像訊號處理裝置之功能模組圖； [0066] 圖2為較佳實施例之圖像訊號處理方法之流程圖； [0067] 圖3為較佳實施例之圖像訊號處理方法之原理示意圖； [0068] 圖4為圖2所示之流程圖之一個子流程圖； [0069] 圖5為圖2所示之流程圖之另一個子流程圖；及 [0070] 圖6為攝像裝置之影像感測器產生之圖像之一種灰階分佈 圖。 【主要元件符號說明】 [0071] 圖像處理裝置：100 [0072] 合併模組：50 [0073] 轉換模組：1〇 096130562 表單編號A0101 第11頁/共20頁 0993265692-0 1334304 099年07月23日按正替換頁 [0074] 存儲模組：60 [0075] 複製模組：20 [0076] 影像感測器：200 [0077] 反轉模組：30 [0078] 公式：71，72，73，74，75 [0079] 去噪模組：40 [0080] 曲線：99 0993265692-0 096130562 表單編號A0101 第12頁/共20頁Yc = Y mn mn The luminance component of the original image Y is copied to generate a sub-picture Yc. For step S03 (reverse step), the inversion module 30 inverts the sub-image Yc using Equation 73: I = D - Yc mn mn to generate a negative image I. Where D is a natural number representing the highest grayscale value of the secondary image Yc, such as the common 256th order or 4096th order (ie D=255 or D=4095). For step S04 (de-noising step), the denoising module 40 uses the formula 74: A = S04(I) mn mn to remove the noise of the negative image I, and generates a denoised negative image A. Wherein, S04(I) represents the processing of the pixel I by the de-arming module 40. For example, the de-moulding mode mn mn group 40 uses an averaging filter or a median filter to perform average filtering (method) or medium on the negative image I. Value filtering (method). The average filtering (method) or median filtering (method) algorithm is simple and can shorten the image processing time. Form No. A0101 Page 8 / Total 20 Pages 0993265692-0 1334304 [0046] [0049] [0052] [0052] Alternatively, a modified averaging filter may be employed or The median filter performs step S04 (de-noising step) to improve the denoising effect of the negative image. Referring to FIG. 4, preferably, step s〇4 (denoising step) of the embodiment includes the following sub-steps: 5041: dividing the negative image I into a complex region; 5042. counting the grayscale average of each region; 5043. determining whether the grayscale value of each pixel exceeds a predetermined range of grayscale average values of the region; and 5044. the grayscale value of the pixel whose replacement grayscale value exceeds a predetermined range of the corresponding grayscale average value is a corresponding grayscale average value.纟 & In this way, it is determined by sub-step S〇43 (decision step) whether or not the pixel is a noised pixel (ie, the grayscale value exceeds the grayscale average of the region where the pixel is averaged), and only the noised pixel is Average filtering. It can filter out the noise and preserve the grayscale feature value of the positive $pixel to avoid the negative image I from being blurred after denoising. The predetermined range can be set according to the user's preference or experience. More preferably, the predetermined range can be determined by the denoising module 4〇. The reference scheme is: sub-step S041 (calculation step) also calculates the gray-scale variance of each region, and the predetermined range corresponding to one pixel depends on the gray-scale variance of the region, such as a double variance or a triple variance. In this way, according to the principle of probability and statistics: the vast majority of information is distributed within the quadratic or triple variance of the mean value of the information, and the grayscale value can exceed the predetermined range of the grayscale mean of the region (two or three times the variance). Pixels are treated as noise pixels. 096130562 Form No. A0101 Page 9/Total 20 Page 0993265692-0 1334304 Correction Replacement Page of July 23, 2009 [0053] For step S05 (combination step), merge module 50 uses Equation 75: [0054] F = S05 ( A) mn mn [0055] A predetermined ratio of the denoised negative image A is incorporated into the luminance component of the original image Y to generate a target image F. Wherein, S05(A) represents the processing of the pixel mn A by the merge module 50. Specifically, in order to preserve the highlight characteristics of the original image Y, the merged mn module 50 should avoid merging the bright pixels of the original image Y. Therefore, S05(A) (merging step) includes the following sub-steps: mn [0056] S051: determining whether the grayscale value of each pixel of the original image Y is smaller than a preset domain value, and agreeing that the grayscale value is smaller than the preset domain The pixel of the value is a dark pixel; [0057] S052: calculating a predetermined ratio value of the grayscale value of each pixel corresponding to each dark portion pixel of the original image Y in the denoising negative image A; and [0058] S053: each predetermined ratio The grayscale value of the value is added to the grayscale value of the dark image corresponding to the original image Y. [0059] The preset field value may be set in advance according to the user's preference or experience. It can be understood that the smaller the number of pixels in the dark portion of the original image Y, the smaller the calculation amount of the merge module 50, and the shorter the image processing time. Therefore, preferably, the preset field value is set such that the number of pixels in the dark portion of the original image Y is smaller than a predetermined number. [0060] The predetermined ratio value may be set in advance according to the user's preference or experience. Specifically, the method further includes the step before the step S05 (merging step): [0061] S0 0 : setting a predetermined ratio value. [0062] A preferred predetermined ratio can be experimentally derived and built into the camera for selection by the user. The predetermined ratio values for this implementation include 10%, 15%, 18%, 20%, and 25%. 096130562 Form No. A0101 Page 10 / Total 20 Page 0993265692-0 1334304 July 23, 2009, the shuttle replacement page [0063] Since the merge module 50 does not merge all the pixels of the original image Y (only dark pixels are merged) . In order to reduce the amount of calculation of the denoising module 40, the image processing time is shortened. Preferably, step S043 (decision step) may only determine whether the grayscale value of each pixel of the negative image I corresponding to the dark portion of the original image Y exceeds a predetermined range of the grayscale average of the region. [0064] In summary, the present invention has indeed met the requirements of the invention patent, and has filed a patent application according to law. However, the above description is only a preferred embodiment of the present invention. Any equivalent modifications or variations made by the person skilled in the art of the present invention should be included in the following claims. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a functional block diagram of an image signal processing apparatus according to a preferred embodiment; [0066] FIG. 2 is a flow chart of a method for processing an image signal according to a preferred embodiment; [0067] 3 is a schematic diagram of a principle of an image signal processing method according to a preferred embodiment; [0068] FIG. 4 is a sub-flowchart of the flowchart shown in FIG. 2; [0069] FIG. Another sub-flow chart; and [0070] FIG. 6 is a gray scale distribution diagram of an image generated by an image sensor of the image pickup device. [Main component symbol description] [0071] Image processing device: 100 [0072] Merge module: 50 [0073] Conversion module: 1〇096130562 Form number A0101 Page 11 / Total 20 pages 0993265692-0 1334304 099 07 Press the positive page on the 23rd of the month [0074] Storage module: 60 [0075] Copy module: 20 [0076] Image sensor: 200 [0077] Inversion module: 30 [0078] Formula: 71, 72, 73,74,75 [0079] Denoising module: 40 [0080] Curve: 99 0993265692-0 096130562 Form number A0101 Page 12 of 20