200910925 九、發明說明： 【發明所屬之技術領域】 本案係為一種判斷影像狀態與其曝光控制之方 法，更特別地，本發明是關於一種判斷影像狀態後可進 一步調整影像之曝光參數之控制方法。 【先前技術】 在般監視糸統中，物件（objects )之辨別一直是 一項重要的課題’其中如何在背光很強的情況下，還能 清楚的辨別物件本身，尤其是人的臉，更是監視系統的 必要功能。 其中200910925 IX. Description of the Invention: [Technical Field] The present invention relates to a method for judging an image state and exposure control thereof. More particularly, the present invention relates to a control method for further adjusting an image exposure parameter after determining an image state. [Prior Art] In the general surveillance system, the identification of objects has always been an important topic. How to clearly identify the object itself, especially the face of the person, especially when the backlight is strong. It is a necessary function of the monitoring system. among them
吳國專利US6,690,424中述及一種背光校正 之方法，其藉由將從整個攝像平面（Emire Image pick叩 Plane )中所求出的平均亮度乘以小於及大於一的係 數以刀別求得用以判斷背景及物體圖像部份（〇bject ene Portion)之免度參考標準，於是可分離出物體圖 像部份’制用前料均亮度與物體®像部份之平均亮 2的土值纟°丨异m修正量。因為是先利用統計圖 表及前述亮度參考標準先分離出物體圖像部份，故與分 區取其權重的方式相較，可聽物體 時所衍生之修正不正麵問題。 置』 :在美國專利US7，167,597中，係利用統計圖表上 該影：ΓίΤ及斜率)，來設定其準則以歸類 “象疋那一類。由於在背光的情況下拍攝物體時，亮 200910925 度統計圖表上會有兩極化的現象，因此反映在統計圖表 的斜率以及頻率/位置上，會有一定的傾向，足以設定 適當的條件作為判斷的準則。因此在滿足上述條件時， 會被適當歸類以做相應的正確修正。此篇專利較偏重於 判斷影像之狀態（如背光影像、於夜間拍攝· · •等狀 態），對於如何修正並未特別著墨。 且在美國公開申請案2006/0171605中，也是必需 先判斷出物件：如人臉，再以物件之内分區（如皮膚區、 頭髮區、皮膚周圍的區域），物件之外以與物件中心之 距離分區，再分別給予各該分區區域一預先設定之權重 以作為調校曝光之用。但其先決條件必先區分出物件區 域及物件以外區域，再區分出物件之内分區區域以及物 件之外分區區域，同時各分區權重為固定值。 因此，本案發明人有鑑於上述習知技術之揭露與其 不足之處，而發明出之「判斷影像狀態與其曝光控制之 方法」，提出一個簡便的判斷影像狀態之方法且可進一 步進行自動調整以執行曝光補償。 【發明内容】 本案之主要目的在於提供一種判斷影像狀態與其 曝光控制之方法，其可提供一合宜的規範，藉以判斷出 是否為完全背光狀態或是完全非背光狀態，再利用增減 最亮及最暗區域計算權重的操作方式，回饋計算並調整 曝光參數，使其依此條件進行曝光補償。 200910925 本案之另一目的在於提供一種判斷影像狀態與其 曝光控制之方法，藉由—自”光補償_在被攝主體 處於背光的情況下’藉由提升主體區域的曝光計算權重 且_降低背景區域的曝光計算權重以提升整體晝面 的冗度’以避免主體區域因背光過暗而失去細節。 』本案之又一目的為一種判斷影像狀態之方法，其 中該影像具有-背景區域與一主體區域，其係包含下 列步驟·（A)將該影像劃分為複數區塊，其中一第一 數目之區塊係位於該背景區域中以及一第二數目之區 塊則落人該主體區域中；（B)偵測料區塊之個別亮 度，藉以計算出該影像之平均亮度；（C)將該背景區 域中^大於該平均亮度之第_倍率之區塊數量加總 為第值，（D)將該主體區域中具有小於該平均亮 度之第二倍率之區塊數量加總為-第二值；以及（E) 將該第-值與該第二值分別與一第一臨界值與一第二 臨界值進行比較，進而判斷該影像之狀態。 卜根據上述構想，當該第一值與該第二值分別大於 ，界值與該第二臨界值時斷該影像係處於 二=全背光狀態。而當該第一值與該第二值分別小於 。玄第臨界值與該第二臨界值時，判斷該影像係處於 一完全非背光狀態。 根據上述構想，該第一倍率係大於1並小於2，較 仏疋”於1.3至1.8之間，該第二倍率係大於〇並小於工， 較佳是介於0.3至0.7之間。 8 200910925 本案之又一目的為提供一種 其中該影像係區分為複數㈣ 心工制之方法， 女々ώ __重區’且該等權魏分別呈 =自的權重與各自的亮度，其包含下列步驟：⑷ 计异該影像之權重平均亮度，藉以 ==於一完全編態;(〇 = 最=度之該權重區的權重，並降低該等權重區 ^=舌之該權重區的權重;以及a)根據該步 驟（C)之榷重調整，重新計算該影像之權 進而調整該曝光參數。 儿没 根據上述構想’該影像係具有-背景區域鱼一主體 區域。因此，該步驟⑻更可包含下列步驟;、㈤广 將该影像劃分為複數區塊，其中該背景區域係且有一第 一數目之區塊，且該主體區域具有—第之 ^叫偵測該等區塊之個別亮L計算㈣ 平^度’（Β3)將該背景區域中具有大於該平均亮度 之第-倍率之區塊數量加總為—第一值；（Β4)將該主 體區域中具有小於該平均亮度之第二倍率之區塊數量 加總為一第二值；以及f r 、a >Λ_ 步m以及⑻）當該第一值與該第二值分 ，大於-第-臨界值與-第二臨界值時，判斷該影像係 處於該完全背光狀態。 根據上述構想’該曝光參數係用於決定一曝光值 (exposure value，EV)，而該曝光值係透過改變一感光 度（ISO)、-快Η速度、-光圈值或是—曝光時間而進 行調整。 200910925 、該根據上述構想，該權重之最大值為4 ’而最小值 為〇,且該等權重區係依照—主體所在而決定各該權重 區的權重。 根據上述構想，該影像係為一監視系統所拍攝之攝 像（Image Pickup )。 本案之又一目的為提供一種影像曝光控制之方法， 其中該影像具有-背景區域和—主體區域，其包含有下 列步驟：U)將該影像劃分為複數個區塊以分別债測 其亮度，其中該背景區域具有-第-數目之區塊，且該 主體區域具一第二數目之區塊；⑻將該等區塊更區分 為一特定數目之權重區，且各該權重區具有各自之權 f ; (c)計算該等權重區之亮度，並依照其各自之權重， 藉以產生該影像之-權重平均亮度；（D)依據該權重 平均亮度，設定該影像之-曝光參數；㈤計算該影像 之:平均亮度；⑺計算該背景區域中具有大於該平 均冗度之第一倍率的區塊數量；（G)計算該主體區域 中具有小於該平均亮度之第二倍率的區塊數量；（H) 分別將該步驟（F)和步驟（G)所計算之區塊數量與 一第一臨界值與一第二臨界值進行比較，進而判斷該影 像之狀悲，（I)因應该影像之狀態而分別調整該等權重 區中具最低亮度之該權重區和該等權重區中具最高亮 度之該權重區的權重；以及（J)因應該步驟（I)之權 重調整’重新計算該影像之權重平均亮度，藉以調整該 曝光參數。 10 200910925 根據上述構想’在該步驟（Η)中，當該第 =::2大於該第-臨界值與該第二臨界值時，到 轉魏中具最低亮度之該權重區的權重，並 牛低4等桑重區中具最高亮度之該權重區的權重。、， 根據上述構想，該步驟（H)中，當該第—值與該 ιΓ值分料於該第—臨界值與該第二臨界值時，判斷 俜為提”耸描舌先 此在該步驟⑴ ^為“該相重區中具最低亮度之該權重區的權重， 並降低該等權重區中具最高亮度之該權重區的權重。 根，上述構想，該平均亮度係為將該㈣塊所伯測 的免度相加後’再除以該等區塊之數量。 3上述構想’重平均亮度係為該等區塊所债 二旦、7C度乘上其所具有的權重後’再除以該等區塊之 數置。A method of backlight correction is described in U.S. Patent No. 6,690,424, which is obtained by multiplying the average brightness obtained from the entire imaging plane (Emire Image pick叩Plane) by a coefficient smaller than or greater than one. The reference standard for judging the background and the image part of the object (〇bject ene Portion), so that the image portion of the object can be separated from the average brightness of the front material and the image of the object image portion. The value 纟° is different from the m correction amount. Since the image portion of the object is first separated by using the statistical map and the aforementioned brightness reference standard, the correction derived from the audible object is not a positive problem compared with the manner in which the partition takes its weight. In US Pat. No. 7,167,597, the use of statistical images on the image: ΓίΤ and slope) to set its criteria to be classified as “like the symbol. Because the object is illuminated in the case of backlight, bright 200910925 degrees There will be a phenomenon of polarization on the statistical chart, so it will be reflected in the slope and frequency/position of the statistical chart. There will be a certain tendency to set the appropriate conditions as the criterion for judgment. Therefore, when the above conditions are met, it will be properly returned. The class is corrected accordingly. This patent is more focused on judging the state of the image (such as backlight image, shooting at night, etc.), and there is no special ink for how to correct it. And in the US open application 2006/0171605 In the middle, it is also necessary to first determine the object: such as the face, and then within the object partition (such as the skin area, the hair area, the area around the skin), the object is divided by the distance from the center of the object, and then given to each of the partitions A pre-set weight for the area to be used as a calibration exposure. However, the prerequisites must first distinguish between the object area and the area outside the object. The partition area within the object and the partition area outside the object are separated, and the weight of each partition is a fixed value. Therefore, the inventor of the present invention has invented the image state and its exposure in view of the above-mentioned prior art disclosure and its inadequacies. The method of control proposes a simple method of judging the state of the image and can further perform automatic adjustment to perform exposure compensation. SUMMARY OF THE INVENTION The main purpose of the present invention is to provide a method for judging the state of an image and its exposure control, which can provide a suitable specification for judging whether it is a full backlight state or a completely non-backlight state, and then using the increase and decrease of the brightest The darkest area calculates the weight operation mode, feedback calculation and adjusts the exposure parameters to make exposure compensation according to this condition. 200910925 Another object of the present invention is to provide a method for judging the state of an image and its exposure control, by - "light compensation _ in the case where the subject is in backlight" by calculating the weight of the exposure of the body region and reducing the background area The exposure calculates the weight to increase the redundancy of the overall surface to avoid the loss of detail in the body area due to the backlight being too dark. Another object of the present invention is a method for determining the state of an image, wherein the image has a background area and a body area. , the method comprising the following steps: (A) dividing the image into a plurality of blocks, wherein a first number of blocks are located in the background area and a second number of blocks are in the body area; B) detecting the individual brightness of the material block, thereby calculating the average brightness of the image; (C) adding the total number of blocks in the background area that is greater than the _ magnification of the average brightness to the first value, (D) Adding a total number of blocks in the body region having a second magnification smaller than the average brightness to a second value; and (E) respectively separating the first value and the second value from a first threshold Comparing with a second threshold value, and further determining the state of the image. According to the above concept, when the first value and the second value are respectively greater than, the boundary value and the second threshold value are interrupted, the image system is at two= a full backlight state, and when the first value and the second value are respectively less than a threshold value and a second threshold value, determining that the image system is in a completely non-backlight state. According to the above concept, the first magnification system If it is greater than 1 and less than 2, it is between 1.3 and 1.8, and the second magnification is greater than 〇 and less than work, preferably between 0.3 and 0.7. 8 200910925 Another object of the present invention is to provide a method in which the image is divided into a plurality (four) mental system, and the female __heavy area' and the weights respectively represent the weights of the self and the respective brightnesses, which include the following Steps: (4) Calculate the weighted average brightness of the image, by == in a complete mode; (〇 = most = the weight of the weight zone, and reduce the weight of the weight zone of the weight zone; And a) re-calculating the image according to the re-adjustment of the step (C) to adjust the exposure parameter. According to the above concept, the image has a body area of the background area. Therefore, the step (8) may further include the following steps: (5) widely dividing the image into a plurality of blocks, wherein the background region has a first number of blocks, and the body region has a - The individual brightness of the equal block L is calculated (4) the flatness '(Β3) is the total number of blocks in the background area having the first-magnification ratio greater than the average brightness - the first value; (Β4) in the body area The number of blocks having a second magnification smaller than the average brightness is added to a second value; and fr , a > Λ _ step m and (8)) when the first value and the second value are divided, greater than - the first critical When the value is - the second threshold, it is determined that the image is in the full backlight state. According to the above concept, the exposure parameter is used to determine an exposure value (EV), and the exposure value is performed by changing an sensitivity (ISO), a fast speed, an aperture value, or an exposure time. Adjustment. According to the above concept, the weight has a maximum value of 4 ′ and a minimum value of 〇, and the weight zones determine the weights of the weight zones according to the location of the body. According to the above concept, the image is an image pickup (Image Pickup) taken by a monitoring system. Another object of the present invention is to provide a method for image exposure control, wherein the image has a background area and a body area, and the method includes the following steps: U) dividing the image into a plurality of blocks to separately measure the brightness thereof. Wherein the background area has a -number-number of blocks, and the body area has a second number of blocks; (8) the blocks are further divided into a specific number of weighted areas, and each of the weighted areas has its own (c) calculating the brightness of the weight zones and generating the weight-weighted average brightness of the image according to their respective weights; (D) setting the image-exposure parameter according to the weighted average brightness; (5) calculating (7) calculating a number of blocks in the background area having a first magnification greater than the average redundancy; (G) calculating a number of blocks in the body area having a second magnification less than the average brightness; (H) comparing the number of blocks calculated in step (F) and step (G) with a first threshold value and a second threshold value, respectively, to determine the shape of the image, (I) corresponding image It Adjusting the weight of the weight zone with the lowest brightness in the weight zone and the weight zone of the weight zone with the highest brightness; and (J) recalculating the image according to the weight adjustment of step (I) The weight is averaged to adjust the exposure parameter. 10 200910925 According to the above concept, in the step (Η), when the first =:: 2 is greater than the first critical value and the second critical value, the weight of the weight region having the lowest brightness in the transfer Wei, and The weight of the weight zone with the highest brightness in the sangzhong area of the cow low 4 and so on. According to the above concept, in the step (H), when the first value and the ι value are distributed to the first critical value and the second critical value, it is determined that the 俜 is a mention. (1) ^ is "the weight of the weight zone with the lowest brightness in the phase weight zone, and reduces the weight of the weight zone with the highest brightness in the weight zones. Root, in the above concept, the average brightness is the sum of the exemptions calculated by the (four) block and then divided by the number of the blocks. 3 The above concept 're-average brightness' is the debt of the blocks, and the 7C degrees are multiplied by the weights they have, and then divided by the number of such blocks.
本案之功效與目的’可藉由下列實施方式說明，俾 有更深入之了解。 【實施方式】 將於下文中說明本發明，熟悉本技術者須曉解下文 的t兄明僅係作為例證用，而不用於限制本發明。 "以下針對本案較佳貫施例之―種制影像狀態與其 曝光控制之方法進行描述，但實際架構與所採行之方法 並不必須完全符合描狀架構與枝，熟習本技藝者當 11 200910925 能在不脫離本發日狀實際精神及範圍的情況下，做出種 種變化及修改。 明參閱第-圖’其係為本發明所揭示之—種影像曝 光控制之方法㈣要流程示㈣。本案之影像曝光控制 方法係先將-影像區分為具不同權重之多個權重區（步 驟101) ’其中該等權重區係因應其在該影像之不同相 應位置而各有不關重，且亦各自具有不同亮度，隨後 因應其權重與亮度而可計算出該影像之權f平均亮度 (步驟1G2) ’因應該權重平均亮度而可設^出該影^ 之曝光參數（步驟1()3)，接著判斷該影像之狀態（步 驟間’以確認該影像是否是—完全f光狀態（步驟 105)、一完全非背光狀態（步驟1〇7)或是一不完全 背光狀態（步驟109)，而當在步驟1〇5中4認該影像 是全者光狀態後，則提高該等權重區中具最低亮度之 權重區的權重’並降低該等權重區中具最高亮度之權重 區的權重（步驟1〇6)，因此根據步驟1〇6之權重調整， 重新回到步驟102去重新計算該影像之權重平均亮 度，進而再§周整該曝光參數。而若是在步驟丨中確認 該影像是完全㈣光狀紐’卿低具最低亮度權重區 之權重與知:咼具最高亮度權重區之權重（步驟1〇8)， 隨後依據步驟108所進行之權重區之調整，重新回到步 驟102去重新計算該影像之權重平均亮度，進而再調整 該曝光參數。此外，當被判斷為均不屬於完全背光狀態 (步驟105)與完全非背光狀態（步驟1〇7)，則被視 12 200910925 為是屬於不完全背光狀態（步驟109)，因此不進行任 何的權重調整（步驟11〇)。 其中’該影像係為一監視系統所拍攝之攝像（Image 腸，該曝光參數係用以決定一曝光值（exposure value，EV)，而該曝光值係透過改變一感光度（is〇)、 i·夬門速度、一光圈值或是一曝光時間而進行調整，且 捕重之最大值為4，而最小值為〇，且該等權重區係 依照一主體所在而決定不同權重區的權重。 此外，前述步驟104中之判斷該影像狀態的詳細流 程更可以參考第二圖，首先將該影像劃分為複數區塊 。(步驟20^1) ’其中該影像更具有—背景區域與一主體 鸛謂5塊中之-第-數目之區塊係位於該背 不:5而一第二數目之區塊係位於該主體區域中， =計算該影像之平均亮度（步驟2G2)，再則，計算 域與該主體區域中分別具有特定條件之 。°鬼數（咖和·)(步‘驟203 )，其中nBt是兮 背景區域中大於該平均真产 μ /干句儿度之弟-倍率之區塊個數，而 域巾㈣小於鮮均亮叙第二倍率 之後料咖接著分別與所設定之個 值比較（步驟2〇4)，當*和 # = 步驟2⑹，則_該影像係為-完二 =地06)，反之，若• 二二 心界值時（步驟207)，則判斷該影 。驟208)，再則’若非屬上述兩種情況者 13 200910925 (即nBt大於其設定臨界值而nDk係小於其設定臨界 值，或是nBt小於其設定臨界值❿祖係大於其設定臨 =)’則判斷該影像係屬於—不完全f光狀態（步驟 ^其中，該平均亮度是將該等區塊所偵測到的亮度相 加後’再除以該等區塊之數量，而該第—倍率一般是大 於1亚小於2 ’較佳是介於13至18之間，該第二倍 率係大於〇並小於1，較佳是介於0.3至〇.7之間。 ★ 一而具體實施本案之判斷影像狀態方法，還請再參閱 第’其係為使用本案方法所配置之影像區塊與區域 ，不意圖。被攝影像30可以劃分為35個區塊31，而 每個區塊31均各自有其不同亮度，將每個區塊所偵 測到的亮度加總再除以該料塊之數量即為該被攝影像 平均亮度胃之完全背総態通常是被攝 ^境之背景亮度過亮而主體亮度過暗，因此本案將被攝 影像30分成一背景區域32和一主體區域33，此背景區 域32具有14個區塊3卜主體區域33具有21個區塊 31，並分別计异將背景區域32中具有大於倍平均 焭度之之區塊31數量加總為一第一值（nBt)以及將主 體=33中具有小於0.5倍平均亮度之區塊數量加總 為一第二值（nDk)，當nBt及nDk分別大於所設定之 ε品界值時（在此實施例中係為nDk^4，nBt^6)，即認 疋被攝影像30係為處於一完全背光狀態，反之，當nDk 與nBt同時小於所設定之臨界值時（在此實施例中係為 14 200910925 扮二6 ) ’則§忍定被攝影像3 〇係處於一完全非 背光狀態。 5月再參考第四圖，其係為利用本案方法所配置之權 重區配置與其權重分布之示意圖。為了便於本案實施例 說明’而將位於不同圖式中的被攝影像以不同圖號表 示，惟其被攝影像之實質實為相同，因此被攝影像40 係與第三时之被攝影像完全相同，且被攝影像30 之區塊配置也與被攝影像3G所配置者完全-致，其區 別僅在於被攝影像4〇更可以依照其不同位置而區分為 7個權重區與5種不同權重分布（41、42、43、44、45)， 其中本實施例係設定被攝影像4 〇之最中間的權重區4! 具最高的權重值L0、其次為中間下方之權重區 〇.8、接著是兩側下方之權重區43各為〇6、中間上方 之權重區44$ 〇.4、最後兩側上方之權重區45各為 〇.2,此種權重分配方式係根據—般影像之主體多為中 間偏下為考量基準’亦可以依料同需求去安排不同方 式的權重區與權值配置’惟此等均等變化仍是屬於本案 之保護形式。此外，根據第三圖之敘述，每個區塊Η :具有其不同亮度’因此該權重平均亮度即是被攝影像 〇中之減區塊分狀亮度分別乘上其所具有的權重 (應所在之權重區不同而有所不同）後再除以該等區塊 之數量之數值。依照此權重平约;^ β 攝影像40之曝衫數 "度’即可據此設定被 於是’依照前面敎述，若在第三圖中被攝影像30 15 200910925 被判斷係屬-完全背光狀態時，且假定在第四圖中，被 攝影像40中之中間下方之權重區42具有最低亮度，而 兩側亡方,權重區45具有最高亮度，因此即可據此調 整具最低亮度之權重區的權重和具最高亮度之權重區的 權重，本實施例係設定每次調整之差值為〇1，亦可以採 用0.01 S 0.25 ’因此該第四圖之權重分配即會調整為第 五圖（A)所示，具最低亮度之權重區51之權重變成〇9, 具具最咼売度之權重區52的權重被調整為〇1，因此， 因應上述之權重調整’被攝影像5〇之權重平均亮度即會 減少，相應地’該曝光參數同時也會被調整，透過改變 -感光度（ISO) 快門速度、—光圈值或是一捧光時 間而改變祕光值明加曝光量，進岐善背光的狀態。 反之，若在第三圖中被攝影像3G被判斷係屬一完 全非背光狀態時，則如第五圖（B)所示，將具最低亮 度之權重區51的權重調低為（^和將具最高亮度之權重 區52的權重調高為〇.3,如此—來，權重平均亮度增加， 相應地，_光參數㈣也會被輕藉以改變其曝光值 而減少曝光量’進而改善完全非背光狀態。而若是1餘 的情況了 ’則屬於不完全背光狀態（即nD]^4，nBt 2 6或是nDkd ’ nB⑸），則維持各權重區之權重不 變權重。 因此，透過本案之判斷影像狀態與其曝光控制之方 法’可以提供簡便的判斷與控制方法關應不同的影像 狀態’並以自動調整的方法來進行曝光補償，且此方法 16 200910925 了、彳盾晨進行，直至完全背光狀態或是完全非背光狀態 不再存在為止。 …紅上所述，本案確實可提供一種判斷影像狀態與其 曝光控制之方法，以提供一個自動曝光補償機制，其將 被攝影像⑽_重分布作核據。#被輯為完全背 光情況下時，將最暗的權重區之權重增加，並將最亮的 權^區之權重減少，藉以提升整體晝面的亮度。在完全 非背光情況時，則減少最暗的權重區之權重增加，並增 加最亮的權重區之權重，藉以抑制整體晝面的亮度 其餘的情況下，卿持各權重區之權重不變，此技術無 須增加許多繁複的軟硬體設備，且實施成本極為低廉。 且’本案判斷影像狀態與其曝光控制之方法可以簡單適 用在-般監視純中卻又提供極高的便利性，因此 有效f進產業之進步’本案技術簡單’可運用領域廣 泛’實具產業之價值’爰依法提出發明專利申請。 以上所述係利用較佳實施例詳細說明本發明，而非 限制本發明的範圍，因此熟知此技藝的人士應能明瞭， 適當而作些微的改變與調整，仍將不失本發明之要義所 在，亦不脫離本發明之精神和範圍，故都應視為本發明 的進-步實施狀況。謹請t審查委員明锻 准，是所至禱。 心 本案得由熟習此技術之人士任施匠思而為諸般修 飾，然S不脫本案申凊專利範圍所欲保護者。 17 200910925 【圖式簡單說明】 係為本案較佳實_之—幻彡料光控制之 方法之簡要流程示意圖。 第二圖係為本案較佳實施例之—__⑽態之 方法之間要流程示意圖。 第三圖係為使用本宰方、丰 域之示意圖。 帛方杨配置以彡㈣塊與區 第四圖係為使用本宰方、、參 i婼舌八* i 法所置之權重區配置與 具權重分布之示意圖。 第五圖（A )和（B )公!2丨丨β 4 m _ 分別進行權重調整之示意圖難為因應不同影像狀態而 【主要元件符號說明】 101將影像區分為多個權重區之步驟 102計算權好均亮度之步驟 103設定曝光參數之步驟 104判斷影像狀態之步驟 105是否為完全背光狀態之步驟 106提高具最低亮度權 權重區之權重之步驟隹重與降低具最高亮度 107是否為完全非背光狀態之步驟 降低具最低亮度權重區 權重區之權重之步驟重與^南具最高亮度 109確認為；ί；完全背光狀態之步驟 18 200910925 110不進行任何權重調整之步驟 201將影像劃分為複數區塊之步驟 202計算平均亮度之步驟 203分別計算背景區域和主體區域中之nBt和nDk之 步驟 204分別與其臨界值比較之步驟 205判斷nBt和nDk是否均大於其臨界值之步驟 206決定為完全背光狀態之步驟 207判斷nBt和nDk是否均小於其臨界值之步驟 208決定為完全非背光狀態之步驟 209確認為不完全背光狀態之步驟 30、40、50被攝影像 31區塊 32背景區域 3 3主體區域 41、42、43、44、45 權重區 51具最低亮度之權重區 52具最南党度之權重區 19The efficacy and purpose of this case can be explained by the following implementations, and we will have a deeper understanding. The present invention will be described hereinafter, and those skilled in the art will be aware that the following is intended to be illustrative only and not to limit the invention. "The following describes the method of the image state and the method of exposure control for the preferred embodiment of the case, but the actual architecture and the method of the method do not have to completely conform to the description structure and branches, and the skilled person is 11 200910925 Various changes and modifications can be made without departing from the actual spirit and scope of this issue. Referring to the first-figure, it is a method for image exposure control disclosed in the present invention (4) to show the flow (4). The image exposure control method of the present invention first divides the image into multiple weight zones with different weights (step 101) 'where the weight zones are not related to each other according to their different positions in the image, and Each has a different brightness, and then the weight of the image f can be calculated according to its weight and brightness (step 1G2) 'The exposure parameter can be set according to the weighted average brightness (step 1 () 3) And then determining the state of the image (between steps to confirm whether the image is a full f-light state (step 105), a completely non-backlight state (step 1〇7), or an incomplete backlight state (step 109), And after recognizing that the image is the all-light state in step 1〇5, increasing the weight of the weight zone having the lowest brightness among the weight zones and reducing the weight of the weight zone having the highest brightness among the weight zones (Step 1〇6), so according to the weight adjustment of step 1〇6, it is returned to step 102 to recalculate the weighted average brightness of the image, and then the exposure parameter is further corrected. If the image is confirmed in step 丨Is finished (4) Light New Zealand's weight and knowledge of the lowest brightness weight area: the weight of the highest brightness weight area of the cookware (steps 1〇8), and then return to step 102 according to the adjustment of the weight area performed in step 108. Recalculating the weighted average brightness of the image, and then adjusting the exposure parameter. Further, when it is determined that none of the images belong to the full backlight state (step 105) and the completely non-backlight state (step 1〇7), it is regarded as 12 200910925 It is an incomplete backlight state (step 109), so no weight adjustment is performed (step 11〇). Where 'the image is a camera taken by a surveillance system (Image intestine, the exposure parameter is used to determine an exposure) Exposure value (EV), and the exposure value is adjusted by changing a sensitivity (is 〇), i·夬 speed, an aperture value, or an exposure time, and the maximum value of the weight is 4, The minimum value is 〇, and the weighting areas determine the weights of the different weighting areas according to the location of the main body. In addition, the detailed process of determining the image state in the foregoing step 104 can refer to the second figure. First, the image is divided into a plurality of blocks. (Step 20^1) 'Where the image has more--the background area and a body 5 means that the block of the -number-number is located in the back: 5 The second number of blocks are located in the body area, = calculating the average brightness of the image (step 2G2), and then, the calculation field and the body area respectively have specific conditions. The number of ghosts (caffe and ·) Step 'Step 203 </ RTI>, wherein nBt is the number of blocks in the background area larger than the average true yield μ / dry sentence, and the area towel (4) is smaller than the fresh average bright second rate Compare with the set value respectively (step 2〇4), when * and # = step 2 (6), then _ the image is - complete two = ground 06), and vice versa, if • two or two core values (step 207 ), then judge the shadow. Step 208), and then 'if it is not the above two cases 13 200910925 (ie, nBt is greater than its set threshold and nDk is less than its set threshold, or nBt is less than its set threshold ❿ ancestral system is greater than its setting Pro =) 'Determining that the image belongs to - incomplete f-light state (step ^ where the average brightness is the sum of the brightness detected by the blocks) and then divided by the number of the blocks, and the The magnification is generally greater than 1 and less than 2', preferably between 13 and 18. The second magnification is greater than 〇 and less than 1, preferably between 0.3 and 〇.7. For the method of judging the image state in this case, please refer to the second section, which is the image block and region configured by the method of the present invention. It is not intended. The photographed image 30 can be divided into 35 blocks 31, and each block 31 is divided into 31 blocks 31. Each has its own different brightness, and the total brightness detected by each block is divided by the number of the blocks, which is the average brightness of the imaged image. The complete back state of the stomach is usually the background of the image. The brightness is too bright and the subject brightness is too dark, so the case will be divided into photographic images 30 The background area 32 and a body area 33 having 14 blocks 3 and the body area 33 having 21 blocks 31, and respectively occupying the block 31 having more than the average average degree in the background area 32 The sum of the numbers is a first value (nBt) and the number of blocks having less than 0.5 times the average brightness in the main body = 33 is added to a second value (nDk), when nBt and nDk are respectively greater than the set ε item boundary The value (in this embodiment, nDk^4, nBt^6), that is, the photographic image 30 is in a completely backlight state, and vice versa, when nDk and nBt are simultaneously smaller than the set threshold value (in In this embodiment, it is 14 200910925, and the second is 6) 'The § is determined to be in a completely non-backlit state by the photographic image 3. In May, refer to the fourth figure, which is the weight area configuration configured by the method of the present case. A schematic diagram of the distribution of weights. In order to facilitate the description of the embodiment of the present invention, the photographic images in different patterns are represented by different figure numbers, but the photographic images are substantially the same, so that the photographic images are 40 and the third time. The photographed image is identical and is photographed like 30 The block configuration is also completely different from that of the photographic image 3G, and the difference is that the photographic image 4 can be divided into 7 weight zones and 5 different weight distributions according to their different positions (41, 42, 43, 44, 45), wherein the present embodiment sets the weight area 4 of the middle of the photographed image 4! The highest weight value L0, followed by the weight area 〇.8 below the middle, and then the lower side The weight area 43 is 〇6, the weight area 44$ 〇.4 in the middle upper part, and the weight area 45 above the last two sides are each 〇.2, and the weight distribution method is based on the fact that the main body of the image is mostly lower In order to consider the benchmark 'can also arrange different weights and weights according to the needs of the demand', these equal changes are still the protection form of this case. In addition, according to the description of the third figure, each block Η has its different brightness. Therefore, the average brightness of the weight is multiplied by the weight of the reduced block in the photographic image, respectively. The weighting zones vary and are divided by the number of such blocks. According to this weight, the number of the photographic images 40 can be set accordingly. According to the above description, if the image is taken in the third picture, 30 15 200910925 is judged to be completely-complete. In the backlight state, and assuming that in the fourth figure, the weight area 42 in the middle of the photographed image 40 has the lowest brightness, and the dead side on both sides, the weight area 45 has the highest brightness, so that the minimum brightness can be adjusted accordingly. The weight of the weight area and the weight of the weight area with the highest brightness. In this embodiment, the difference between each adjustment is set to 〇1, and 0.01 S 0.25 ' can also be used. Therefore, the weight distribution of the fourth figure is adjusted to the first As shown in the fifth diagram (A), the weight of the weight area 51 having the lowest brightness becomes 〇9, and the weight of the weight area 52 having the most ambiguity is adjusted to 〇1, and therefore, the photographic image is adjusted in accordance with the above weight. The average brightness of the weight of 5〇 will be reduced, and accordingly the 'exposure parameter will also be adjusted, and the secret light value will be changed by changing the sensitivity (ISO) shutter speed, the aperture value or a holding time. Quantity, into the state of good backlightOn the other hand, if the photographed image 3G is judged to be in a completely non-backlit state in the third figure, as shown in the fifth diagram (B), the weight of the weight region 51 having the lowest brightness is lowered to (^ and The weight of the weight area 52 with the highest brightness is increased to 〇.3, so that the average brightness of the weight is increased, and accordingly, the _light parameter (4) is also borrowed to change the exposure value and reduce the exposure amount, thereby improving the completeness. Non-backlight state. If there is more than one case, then it is incomplete backlight state (ie, nD]^4, nBt 2 6 or nDkd ' nB(5)), then the weights of the weight zones are kept constant weight. Therefore, through this case The method of judging the image state and its exposure control 'can provide a simple judgment and control method corresponding to different image states' and automatically adjust the exposure compensation method, and this method 16 200910925, 彳 shield morning, until completely The backlight state or the completely non-backlight state no longer exists. ... The red case, the case does provide a method for judging the image state and its exposure control to provide an automatic exposure compensation machine. , which will be re-distributed by the photographic image (10)_. When # is compiled as a full backlight, the weight of the darkest weight area is increased, and the weight of the brightest weight area is reduced, thereby improving the overall 昼The brightness of the surface. In the case of completely non-backlighting, the weight of the darkest weight area is increased, and the weight of the brightest weight area is increased, so as to suppress the brightness of the overall surface, in the case of remaining weights, The weight does not change, this technology does not need to add a lot of complicated hardware and software equipment, and the implementation cost is extremely low. And 'the method of judging the image state and its exposure control can be simply applied to the general monitoring but providing high convenience. Therefore, the progress of the industry is effective. 'The technology is simple in this case' can be applied in a wide range of 'real value of the industry'. The invention patent application is filed according to the above. The above description is based on the preferred embodiment, but the invention is not limited thereto. The scope of the art, therefore, those skilled in the art should be able to understand, and make minor changes and adjustments as appropriate, without losing the essence of the invention, and without departing from the scope of the present invention. The spirit and scope of the Ming Dynasty should be regarded as the further implementation of the present invention. I would like to ask the members to examine the foresight and the prayers. The heart of the case must be modified by the people who are familiar with the technology. However, S does not deviate from the scope of the patent application claimed in this case. 17 200910925 [Simple description of the diagram] This is a brief flow diagram of the method of the illusion of light control in this case. The second picture is the case A schematic diagram of the process between the methods of the __(10) state of the preferred embodiment. The third diagram is a schematic diagram of the use of the stagnation and Fengyu. The 帛方杨 configuration uses the 彡(4) block and the fourth map of the district as the use of the slag Schematic diagram of the allocation of weights and the distribution of weights in the 婼 婼 * * * 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 权 权 。 。 The different image states and the main component symbol descriptions 101. The step 102 of dividing the image into a plurality of weight regions, the step 103 of calculating the brightness uniformity, the step 104 of setting the exposure parameters, and the step 104 of determining whether the image state is 105 is the full backlight state. mention The step of lowering the weight of the lowest brightness weight area and the step of lowering the weight of the lowest brightness weight area weighting step with the highest brightness 107 are confirmed as the highest brightness 109; Step 18: full backlight state step 18 200910925 110 does not perform any weight adjustment step 201 to divide the image into complex blocks step 202 to calculate the average brightness step 203 respectively calculate the background region and the body region nBt and nDk step 204 respectively Step 205 of comparing step 205 of determining whether nBt and nDk are greater than their critical values, step 207 of determining whether the nBt and nDk are all in a full backlight state, determining whether nBt and nDk are both less than their critical values, is determined to be a completely non-backlit state step 209. Steps 30, 40, 50 confirmed as incomplete backlight state are photographed 31 block 32 background area 3 3 main body areas 41, 42, 43, 44, 45 weight area 51 has the lowest brightness weight area 52 has the most southern party Weight area 19