201023628 九、發明說明： 【發明所屬之技術領域】 尤指一種全景 影像 產生=明係影像產生方法， 【先前技術】 一般的數位拍攝事晋 機 曰後回 參 參 來舉例說#能拍_定’σ ^目，（以下皆以數位相 憶。但數位相機拍攝的視角景定者曰後 四周的美景全部拍攝在同 關’I法將環餐 憾。 像中’因此常造成許多廷 於疋’漸漸發展出—種可拍 機，所謂全景照片，就-:片之功旎的和 片，然後通過相關軟體拼接 分=攝為多_ _度的景象壓縮:一影類別’它將超 人眼視角平時所能容納，如現的内容遠遠超遇 , 如此、，、°人^來極強的視覺衝擊。 ^ .目機在進行全景拍攝時，大都係將連續拍 攝之影像先個顺私來，當接合雜時，再進^ 像縮小的動作，以加快處理速度，盡看/ 到全景影像。請參考第—圖，該_為習知之全景 實施例之示意圖。如第—圖所示，使用者對目標 1連續拍攝複數張局部影像後於後端接合，而在顯示影 像時，係將接合後的全景影像！依照營幕的可視區域U 來縮放成適當大小’以利使用者觀看。如第二圖所示，可 201023628 視區域11中具有兩張局部影像lu、113，當使用者進^ 放大時，便可從所拍之全景影像i逐漸改變縮小比例來^ 到放大的目的。 $ 】後接，係將原本拍攝的影像縮 =所拍之全景影像！，互動地移動或放大影 =興趣的區域時，—般相機僅是姆地平移並利用既有的 ϊ:ϊΐΓΓ放大，當其放大倍率超過全景影像之原 考第-乂’：：呈現出的結果必然是模糊的影像，請參 考第二圖，由於放大局部影像⑴、113 度時，其中的影像便產生模糊的視覺效果，因斤 一般使时的f求。 滿足 專傻in決低解析度的缺點，另外提出採用較高解析的 景影像’於接合完畢後，再將其縮放至適 =、來顯示於螢幕。如此雖然於放大影像時，可以從以 南解析之影像擷取資剩·亦 然這種方式需要耗費大量得:像高、=像細節， 間亦會讓使用者感到不適像接口時間’過長的等待時 【發明内容】 於先mm失真之影像為目標，本案發明人有感 Γ時間過長或成本偏高等問題’故 析度不ί時==?像來接合成全景影像’若解 更使用原始大小之局部影像來接合，期能提 201023628 =高的影像接合效率，並能互動地顯示更高解析的全景 能解在提供-種全景影像產生方法，俾 位拍=生方法’係適用於-數 步驟如下:i:，；二=景影像產生方法之執行 每-局^禮始，U攝複數個局部影像；其次，將 -接d以產生複數個低解析度局部影像;進行 •景-全景， 景影像之妨超 預攻值，並確認該全 r 之座 ==:所對應之相對解析度較二= G式:力有：:㈣其他目的及優點，將在後續的= 實施方式】 合 在進行全景拍攝完後，大都會 ，以節省後置處理時間，奋=尺寸之影像來接 時，其所呈現出 田使用者邊移動影像邊放大 係而顯得更模糊，二、解析度較低的關 明提出-種較經濟的多解析處來呈=細= 201023628 分，此多解析之全景接合，既可保持原有的接合效 能互動地顯示更高解析的全景影像。 ，並 .切犯顯不更尚解析的全景影像。 — 像依 本發明主要技術特徵在於利用原本拍攝之高解 需求進行部份的接合已呈現影像的細節成分，X影 僅提出必要之動作流程，然而，熟悉該項技藝者得知下= 了以下所提及之流程步驟，數位影像處理方式當: 他的必要步驟，因此，不應以本實施例揭露者為限^括其 ❹ m =，請參考第四圖，該圖係為本發明所揭示 2之一具體實施例之示意圖。如第四圖所示，使用f 旦藉=數位拍攝裝置對目標之—全景連續拍攝複數 馳局部影像縮小後來組合成寬廣的= c。#由相關軟體接合後的全景影像2可依昭= 縮放成適當大小，以利使用者觀看 = 免在拍攝二 部影像接合成較無縫Lit區域來將低解析度局 儲存大每-拍攝之局部影像會^ 當使用者對可 ^仃‘小以利快速接合。因此’ :區域^中的部份 8 3；^：^ 201023628 示的部份進行縫合，利用對， 加快縫合速度。 J用對小部份影像進行處理來 最後，請請參閱第六圖， 景影像產生方式之—具 Μ係為本發明所揭示之全 相闕之局部影像，請同時參二：之步驟流裎圖。按，其t c像產生方式 2五圖。如第六 _)’並對目標全景拍攝複數全景拍攝模式（步驟 再將其儲存於記憶體中（步驟s f °卩影像（步驟S603)， 像縮小成適當比例來屋 ，接著，對每一局部影 S607);為了日後需要切換顯固=解析度局部影像（步驟 記錄低解析度局部影像轉換大小的局部影像，仍需 驟漏），其中，該轉換參數包象之一轉換參數（步 等資訊。 座‘、角度、縮放比例 接著，使用者可任意瀏覽、 低解析度局部影像（步驟s_)警旦^或拉近地觀賞 統會判斷目前可視區域21中顯干象的過程中，系 以上之低解析度局部影像(步心Ϊ否 影像；若步驟S611的判斷為是，便進行—接合程序，、以將 =視區域21巾的料低解減局部雜接合成全景影 像2(步驟S613)，其中當然包括了調整相鄰低解析度局部 影像的角度’以及比對重疊區域之步驟。 再來，判斷目前瀏覽之影像的放大倍率是否超過一預 設值（S615)，其中該預設值係為全景影像2的放大倍率應 有之最低預设值，右否則便繼續劇覽影像；若步驟j 5 201023628 的判斷為是，表示已超二、 最大解析度’則利用先前儲析度局部影像的 區域，的所對應之局部影像==，，可視 而在轉換影像之前，系絲合二力驟則7)。 像是否已存在快取暫存單元r 清所欲顯示之轉換影 顯示存放於快取暫存單 職中讀取並 開始進行轉換影像的動像（_，·若否，則 33於可視區域21中的部^#棘該等局部影像3卜 以符合目前的放大倍#(s 部影像 部份之局部影像31、3S ϋ的角度、亮度（卿），比對 域來接合部;之:部 最後，將部份之局部影像轉換完 ❿ (步驟H 彡像，並將㈣祕錄暫存單元 之邱^八 置衫像中放大部份通常係為使用者有興趣 暫二=再觀#的機會較高，周此可將轉換後的影像 暫存於快取中’以利日後節省轉換影像的時間。 藉由以上具體實例詳述，#可 = 產生方法係當進行全景拍攝完後，先按= 影像縮小來進行接合，且原有之較高解析 像亦-併儲存於記憶卡，其中並把接合所需之座 不轉換相關參數儲存於相對的影像中（如標頭檔裡），當使 =者互動地移動影像並放大時，若其放大後之解析度小於 原全景影像之解析度時，便從原全景影像減少縮小比例來 201023628 達成影像放大的效果，然當放大倍率 一 系統無法利用既有的全景影像來縮小達成日吏 先刖儲存之局部影像，擷取其相對之可視區域的部份 實際顯不之倍率來縮小影像，並依據先前所儲存之相關= ，轉f參數來將部份的局部影像轉換後，貼合於可見^ 參 ❹ 塊㈣區域)的局部影像來互二處理-小 析度，得以節省顯示全景影像的時間。又，cm解 近期轉換過之影像又須再次轉換像，以利節省 *圖上=用具體實施例之詳細說明 以下述之申請專利範圍為準，任何孰斤有範圍應 :之領域内，可輕易思及之變化或；::可：，本發 案所界定之專利範圍。 涵盍在以下本 【圖式簡單說明】 第一圖係為習知之全景影像之一具體實 =二圖係為習知之局部影像之-具 :芯; 圖 第三圖軸W局部败意 意圖第四以圖^系為本發明所揭示之全景影像之一具體實施例之示 意圖第為本發明所揭示之局 201023628 第六圖係為本發明所之全景影像產生方法之一具體實施 例之步驟流程圖。 【主要元件符號說明】 習知： I :全景影像 II :可視區域 111、113 :局部影像 本發明： 2:全景影像 21 :可視區域 31、33 :局部影像 S601〜S633 :各個步驟流程201023628 IX. Invention Description: [Technical field of invention] Especially a panoramic image generation = Ming image generation method, [Prior Art] General digital shooting, Jinji, after returning to the reference, for example, #拍拍_定'σ^目, (The following are all digitally recalled. But the panoramic view taken by the digital camera is all around in the same scene. The I method will be a ring meal. The image is often caused by many ' Gradually developed - a kind of camera, the so-called panoramic photo, on-: the film and the film, and then through the relevant software stitching = shooting as more _ _ degree of scene compression: a shadow category 'it will superhuman eye The angle of view can be accommodated in peacetime, as the current content is far superior, so, and the people have a strong visual impact. ^. When the camera is shooting in a panoramic view, most of the images will be taken first and foremost. Come, when the joint is mixed, then zoom in and out to speed up the processing and look at the panoramic image. Please refer to the figure, which is a schematic diagram of the conventional panoramic embodiment. , the user is connected to the target 1 After capturing a plurality of partial images, they are joined at the rear end, and when the image is displayed, the combined panoramic image is zoomed to an appropriate size according to the visible area U of the camp to facilitate viewing by the user. As shown in the second figure. , 201023628 The viewing area 11 has two partial images lu, 113. When the user zooms in, the zooming ratio can be gradually changed from the captured panoramic image i to the purpose of zooming in. $ 】 Reduce the original captured image to the captured panoramic image! When interactively moving or magnifying the shadow = area of interest, the camera is only panning and using the existing ϊ: ϊΐΓΓ zoom, when its magnification exceeds the panorama The original image of the image-乂':: The result is necessarily a blurred image. Please refer to the second picture. Since the partial image (1) and 113 degrees are enlarged, the image will produce a blurred visual effect. When the f is satisfied, it satisfies the shortcomings of stupidity and low resolution. In addition, it is proposed to use the higher resolution scene image. After the joint is completed, it is scaled to the appropriate = to display on the screen. In the case of images, you can extract the remaining images from the south. This method also requires a lot of cost: like high, = like details, and the user will feel uncomfortable when the interface time is too long. SUMMARY OF THE INVENTION The image of the first mm distortion is the target, and the inventor of the present invention has a problem that the time is too long or the cost is too high, so when the resolution is not good, the image is joined to a panoramic image, and the original size is used. Partial image to join, can improve 201023628 = high image bonding efficiency, and can interactively display higher resolution panoramic energy solution in the provision of a kind of panoramic image generation method, the position of the shot = the method of life is applicable to the - number of steps As follows: i:,; two = scene image generation method execution per-office ^ ceremony, U to capture a plurality of partial images; secondly, will be connected - to generate a plurality of low-resolution partial images; The image of the scene may exceed the pre-attack value, and confirm that the full r seat ==: the corresponding relative resolution is lower than the second = G type: force has:: (d) other purposes and advantages, will be in the subsequent = implementation After the panoramic shooting, the metropolis, to the festival After the province's post-processing time, when the image of the image of the image is connected, the user of the field appears to be more blurred when the image is moved and the image is enlarged. Second, the lower resolution is proposed - a more economical multi-analysis At the same time = fine = 201023628 points, this multi-resolution panoramic joint can maintain the original joint performance interactively to display higher resolution panoramic images. And cut the panoramic image that is not even more analytical. - The main technical feature of the present invention is that the partial stitching of the image has been presented using the high resolution requirements of the original shooting, and the X shadow only proposes the necessary motion flow. However, those skilled in the art know that the following = The process steps mentioned, the digital image processing method is as follows: his necessary steps, therefore, should not be limited to the disclosure of this embodiment, m =, please refer to the fourth figure, which is the invention A schematic diagram of one embodiment of the disclosure 2 is disclosed. As shown in the fourth figure, the target-span-continuous shot is used to reduce the partial image size and then combine it into a wide = c. #The panoramic image 2 joined by the relevant software can be scaled to the appropriate size for the user to view = Free to capture the two images into a more seamless Lit area to store the low resolution office. The partial image will be ^ when the user is ready to use the small to facilitate quick engagement. Therefore, the part of the area: 8:; ^: ^ 201023628 is stitched, and the stitching speed is accelerated by the pair. J uses a small part of the image processing. Finally, please refer to the sixth picture. The scene image generation method is the partial image of the full phase of the invention disclosed in the present invention. Please also refer to the second step: . Press, its t c image generation mode 2 five maps. For example, the sixth _)' and the plurality of panoramic shooting modes are taken for the target panorama (steps are stored in the memory (step sf ° 卩 image (step S603), the image is reduced to an appropriate ratio to the house, and then, for each part Shadow S607); in order to switch the display=resolution partial image in the future (step recording the partial image of the low-resolution partial image conversion size, still need to leak), wherein the conversion parameter package is one of the conversion parameters (step information, etc.) Block ', angle, zoom ratio, then the user can browse, low-resolution partial image (step s_), or close to the viewing system to determine the current visible image in the visible area 21, above The low-resolution partial image (step-by-step image); if the determination in step S611 is YES, the bonding process is performed, and the material of the view area 21 is low-decomposed and partially merged into the panoramic image 2 (step S613). ), of course, including the step of adjusting the angle of the adjacent low-resolution partial image and the step of comparing the overlapping regions. Then, determining whether the magnification of the currently viewed image exceeds a preset value S615), wherein the preset value is the lowest preset value of the magnification of the panoramic image 2, and the right continues to the drama image; if the judgment of the step j 5 201023628 is YES, it indicates that the maximum resolution is exceeded. 'The corresponding partial image of the area where the partial image of the previous resolution is used is ==, and it can be seen that before the image is converted, the stitching force is followed by 7). If there is already a cached temporary storage unit r The conversion image to be displayed is displayed in the cached temporary job and read and starts to convert the image of the image (_, if not, then the portion of the image in the visible area 21 In order to conform to the current magnification # (the partial image of the image portion of the s portion, the angle of the 3S ϋ, the brightness (clear), the alignment field to the joint; the: the end, the partial partial image is converted ( Step H 彡 , 并将 , , , , , 秘 秘 秘 秘 秘 秘 秘 秘 秘 ^ ^ ^ 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大 放大Temporarily stored in the cache to save time in converting images in the future. Detailed description of the specific example, #可= The generation method is to perform the panoramic shooting, first press = image reduction to join, and the original higher resolution image is also - stored in the memory card, and the bonding required The parameters of the non-conversion are stored in the relative image (such as the header file). When the image is moved and magnified interactively, if the resolution after amplification is smaller than the resolution of the original panoramic image, then the original The panoramic image reduces the reduction ratio to 201023628 to achieve the image enlargement effect. However, when the magnification ratio system cannot use the existing panoramic image to reduce the partial image stored in the future, the part of the relative visible area is actually displayed. The magnification is not reduced to reduce the image, and the partial image is converted according to the previously stored correlation =, and the partial image is converted to the local image of the visible (4) region. Degrees save time in displaying panoramic images. In addition, the solution of the recently converted image of the cm solution has to be converted again to save the image. * The detailed description of the specific embodiment is based on the following patent application scope, and any scope of the product should be: Easily think about changes or::: Can: The scope of patents defined in this case. The following is a brief description of the schematic image. The first image is one of the conventional panoramic images. The second image is a partial image of the conventional image: with the core: the third image of the axis W is partially defeated. 4 is a schematic diagram of a specific embodiment of a panoramic image disclosed in the present invention. The present invention is disclosed in the present invention. 201023628. The sixth figure is a flow chart of a specific embodiment of a panoramic image generating method of the present invention. Figure. [Main component symbol description] Convention: I: panoramic image II: visible area 111, 113: partial image The present invention: 2: panoramic image 21: visible area 31, 33: partial image S601~S633: various step processes
1212