12725521272552
〔發明所屬之技術領域〕、4 本案為一種將浮水印植入影像中 (Spatial DomainU ,將分割的浮水 =,尤指在空間域 列(Video Sequence)的圖框fpr Λ 刀散植入視訊序 η lL- 、frame)中 〇 〔先前技術〕 隨 媒體儲 位媒體 的風險 者的智 浮 記隱藏 位媒體 位媒體 〔發明 本 去,係 令散植 !fl序列 貧訊。 本 友法’ IfL 9 ίλ 〔發明 著電子 存,以 得以快 。因此 慧財產 水印技 至被保 中植入 中取出 目的〕 案的目 在空間 入視tfL 在經過 案的另 係在視 減少f子 内容] 化社會的來臨,女旦你 利保存與傳播。然=,、钿統媒體被轉換為數位 ig Μ P ^ Ur…、 、、際網路的普及,使數 迷稷I與政佈,相對的, 仗数 ’提供數位媒體的版權保護二:未經授權使用 權,乃是刻不容緩的二制,以保護創作 術(watermarking) ’ 是將可处 A 護資料中的技術。因此,在權所I的標 浮水印資訊,達到版權保護:作義::便可由數 的,即在於建立一種將浮水印植 域(sPaual DomainXh,將分割的=中= 序列(Video Se(luence)的圖框(F子水P衫像 頻域轉換操作* ’仍能取出可供識:e的浮而水視印 ^的,在於建立一種將浮水印植 訊序列中的單張圖框,只植入部分浮:中: 原始影像的破壞。 〇矛水印-貝[Technical field to which the invention pertains], 4 This case is a method of implanting a watermark into an image (Spatial DomainU, which divides the floating water =, especially in the frame of the spatial sequence (Frequency Sequence), the image of the fpr 散 knife is scattered into the video sequence. η lL- , frame ) 中〇 [Prior Art] With the media storage media risk of the wisdom of the memory hidden media media [invented this way, the system is scattered! This friend's law, IfL 9 ίλ [invented the electronic deposit, so that it can be fast. Therefore, the Huishui watermarking technology is taken out of the insured object. The purpose of the case is to look into the tfL in the space. The other is to reduce the content of the sub-content. The society is coming, and the woman is saved and spread. However, the media has been converted into digital ig Μ P ^ Ur..., , and the popularity of the Internet, so that the number of confused I and the political cloth, in contrast, the number of 'copyright copyright protection for digital media II: No Authorized use rights are the two systems that cannot be delayed, to protect the watermarking' is the technology that will be available in the A protection materials. Therefore, the standard watermark information in the right of the I achieve copyright protection: the meaning:: can be counted, that is, to establish a watermarking domain (sPaual DomainXh, will be segmented = medium = sequence (Video Se (luence) The frame of the F-water P-shirt image frequency domain conversion operation* can still be taken out for identification: the floating and water-like printing of e is to establish a single frame in the watermarking sequence. Only implant part of the float: Medium: Destruction of the original image.
第4頁 1272552 五、發明說明(2) 為達上述目的,本案提出· 一種將浮水印(W a t e r m a r k )植 入影像的方法,其步驟為:提供一浮水印及複數個影像;分 割該浮水印為複數個部份浮水印;以及植入該複數個部分浮 水印至該複數個影像,使該複數個影像中包含該浮水印的資 訊。 如所述之將浮水印植入影像的方法,其中該浮水印可為 一單色影像、一灰階影像或一彩色影像。 如所述之將浮水印植入影像的方法,其中該複數個部分 浮水印係分散地植入該複數個影像。 如所述之將浮水印植入影像的方法,其中該複數個部分 浮水印可在空間域(S p a t i a 1 D 〇 m a i η )植入該複數個影像。 如所述之將浮水印植入影像的方法,其中該複數個部分 浮水印可在頻率域(F r e q u e n c y D 〇 m a i η )植入該複數個影像。 如所述之將浮水印植入影像的方法,其中該複數個影像 可為一未壓縮之視訊序列(V i d e 〇 S e q u e n c e )的複數個圖框 (Frame)0 如所述之將浮水印植入影像的方法,其中該複數個影像 信號可為一壓縮之視訊序列的複數個圖框(Frame)。 如所述之將浮水印植入影像的方法,其中該壓縮之視訊 序列可為Mpeg2或Mpeg4視訊序列。Page 4 1272552 V. Invention Description (2) In order to achieve the above purpose, the present invention proposes a method for implanting a watermark (W atermark) into an image, the steps of which are: providing a watermark and a plurality of images; dividing the watermark And a plurality of partial watermarks are implanted; and the plurality of partial watermarks are implanted into the plurality of images, so that the plurality of images include the information of the watermark. A method of implanting a watermark into an image as described above, wherein the watermark can be a monochrome image, a grayscale image, or a color image. A method of implanting a watermark into an image as described above, wherein the plurality of partial watermarks are discretely implanted into the plurality of images. A method of implanting a watermark into an image as described above, wherein the plurality of partial watermarks can implant the plurality of images in a spatial domain (S p a t i a 1 D 〇 m a i η ). The method of implanting a watermark into an image as described above, wherein the plurality of partial watermarks can implant the plurality of images in a frequency domain (F r e q u e n c y D 〇 m a i η ). The method for implanting a watermark into an image, wherein the plurality of images may be a plurality of frames of an uncompressed video sequence (V ide 〇 Sequence), as described above, The method for entering an image, wherein the plurality of image signals can be a plurality of frames of a compressed video sequence. A method of implanting a watermark into an image as described, wherein the compressed video sequence can be an Mpeg2 or Mpeg4 video sequence.
像 影入 植 印水 浮 將之 例 施 實 佳 1272552 五、發明說明(3) 中的方法。如圖所示,在、空間·域下,於視訊序列VS( Video Sequence)的每個圖框(Frame)中植入兩行浮水印WM的部分資 訊,即視訊序列VS的第一個圖框F 1之最上一行TL 1及最下一 行BL1分別植入浮水印WM的第0行L0和第1行L1,···,圖框 Fn/2的TLn/2及BLn/2分別植入浮水印WM的最後兩行Ln-2和 Ln-1,而圖框Fn/2 + l的TLn/2 + l及BLn/2 + l則重新攫取浮水印 WM的L0和L1植入,直到視訊序列VS的最後一個圖框Fm都植入 兩行浮水印WM的部分資訊。其中浮水印WM為單色的點陣圖 (Bitmap),而視訊序列VS則為YU V4 2 0格式。 第二圖為浮水印植入圖框的示意圖,在此舉浮水印龍的 第1行L0植入圖框F 1的最上一行TL1做說明,餘類推。如圖所 示,L0的每一點係因應浮水印WM與視訊序列VS中圖框的尺寸 比例,而展成多點植入TL卜在此若以88*72的浮水印WM植入 C I F ( 3 5 2 * 2 8 8 )的視訊序列V S裡為例,則l 〇的每一點會展成4 點植入TL1中。不過,由於浮水印〇為單色的點陣圖 (Bi tmap),每一點只佔1位元(Bi t)大小,而視訊序列vs的點 為1位兀組(Byte)。因此,浮水印醫的點只能植入位元組其 中的1 b 11。第二圖所示的,即為L 〇的每一點展成4點植入 TL1各點位元組_中的bit_6i示意圖。第四圖則為實際的操作 結果,圖中顯示視訊序列vs植入浮水印WM( ijm)後,再經 Mpeg4壓縮及解壓縮處理的結果,以及從中取出的浮水印 WMout ’其中浮水印醫的點係植入視訊序列vS各點的bi t —6。 當然丄浮水印的點也可植入視訊序列各點的其他位元, 不過植入高位兀會讓畫面上下緣產生顫抖的現象,但粹取出The method in the invention (3) is as shown in the example of the invention. As shown in the figure, in the space and domain, part of the information of the two rows of watermark WM is implanted in each frame of the video sequence VS (Video Sequence), that is, the first frame of the video sequence VS The uppermost row TL 1 and the lowermost row BL1 of F 1 are respectively implanted in the 0th line L0 and the 1st line L1 of the watermark WM, and TLn/2 and BLn/2 of the frame Fn/2 are respectively implanted in the floating The last two lines of the watermark WM, Ln-2 and Ln-1, and the TLn/2 + l and BLn/2 + l of the frame Fn/2 + l re-fetch the L0 and L1 of the watermark WM until the video sequence The last frame of VS, Fm, is embedded with partial information of two rows of watermark WM. The watermark WM is a monochrome bitmap (Bitmap), and the video sequence VS is a YU V4 2 0 format. The second figure is a schematic diagram of the watermarking implanted frame, in which the first row of the watermarking dragon L0 is implanted in the uppermost row TL1 of the frame F1 for illustration, and so on. As shown in the figure, each point of L0 is integrated into the multi-point implant TL according to the size ratio of the watermark WM and the frame in the video sequence VS. Here, if the watermark WM of 88*72 is implanted into the CIF (3) For example, in the video sequence VS of 5 2 * 2 8 8 ), each point of l 会 is expanded into 4 points and implanted in TL1. However, since the watermark is a monochrome bitmap, each point occupies only 1 bit (Bi t), and the video sequence vs is 1 bit Byte. Therefore, the point of the watermark doctor can only be implanted in 1 b 11 of the byte. In the second figure, each point of L 〇 is shown as a 4-point map of bit_6i implanted in the byte _ of each point of TL1. The fourth picture is the actual operation result. The figure shows the result of the video sequence vs implanted the watermark WM ( ijm), then the Mpeg4 compression and decompression process, and the watermark WMout taken from it. The point is implanted into the bi t-6 of each point of the video sequence vS. Of course, the point of the watermark can also be implanted in other bits of the video sequence, but the implantation of the high position will cause the upper and lower edges of the picture to tremble, but the detachment
1272552 五、發明說明(4) 來的浮水印最清晰;反之則較·為模糊。另外,浮水印除了使 用單色影像,也可為灰階影像或彩色影像。而為了達到防偽 的效果,也可以將視訊序列中的圖框縮小,合併版權資訊作 為浮水印。 綜上所述,本案係為一種將浮水印植入影像的方法,除 了較佳實施例所揭露的實施方式外,只要係將分割的浮水印 資訊分散植入複數個影像中,使單一影像中只包含部分的浮 水印資訊,而複數個影像中則包含完整可粹取的浮水印資 訊,即為本案的特徵。其中,本案浮水印植入影像的方法可 在空間域(Spatial Domain)或頻率域(Frequency Domain)中 進行。被植入浮水印的影像,除了未壓縮之視訊序列外,也 可為Mpeg2或Mpeg4等壓縮格式之視訊。 本案所揭露之技術,得由熟習本技術人士據以實施,而 其前所未有之作法亦具備專利性,爰依法提出專利之申請。 惟上述之實施例尚不足以涵蓋本案所欲保護之專利範圍,因 此,提出申請專利範圍如附。1272552 V. Inventive Note (4) The watermark is the clearest; otherwise, it is more vague. In addition, the watermark can be a grayscale image or a color image in addition to a monochrome image. In order to achieve the anti-counterfeiting effect, the frame in the video sequence can also be reduced, and the copyright information is combined as a watermark. In summary, the present invention is a method for implanting a watermark into an image. In addition to the embodiment disclosed in the preferred embodiment, the split watermark information is dispersed into a plurality of images to make a single image. It only contains part of the watermark information, and the multiple images contain the complete watermark information, which is the feature of this case. The method for depositing images by the watermark in the present case can be performed in a spatial domain or a frequency domain. The image that is implanted with the watermark can be a video of a compressed format such as Mpeg2 or Mpeg4 in addition to the uncompressed video sequence. The technology disclosed in this case can be implemented by a person familiar with the technology, and its unprecedented practice is also patentable, and the application for patent is filed according to law. However, the above embodiments are not sufficient to cover the scope of patents to be protected in this case. Therefore, the scope of the patent application is attached.
1272552 圖式簡單說明 〔圖示簡單說明〕 , _ 本案得藉由下列圖示及詳細說明,俾得一更深入之瞭 解: - 第一圖:本案較佳實施例之將浮水印植入影像中的方 法。 第二圖:浮水印植入圖框的示意圖。 第三圖:浮水印的點植入圖框各點之b i t - 6的示意圖。 第四圖:本案較佳實施例的實際操作結果。 圖不主要元件之圖號如下· WM:浮水印(Watermark) VS:視訊序列(Video Sequence) F 1,…,Fn/2,Fn/2 + 1,…,F m·圖框(Frame) LO, LI,···,Ln-2, Ln-1 :浮水印的每行資訊 TL1,…,TLn/2,TLn/2+l,…:圖框的最上一行 BL1,…,BLn/2,BLn/2+l,…:圖框的最下一行1272552 Simple description of the diagram [simple description of the diagram], _ This case can be further understood by the following diagram and detailed description: - First picture: The watermark is implanted into the image in the preferred embodiment of the present invention. Methods. Figure 2: Schematic diagram of the watermark implanted frame. Figure 3: Schematic diagram of b i t - 6 at each point of the watermarked point of the watermark. Figure 4: Actual operation results of the preferred embodiment of the present invention. The figure of the main components is as follows: WM: Watermark VS: Video Sequence F 1,...,Fn/2,Fn/2 + 1,...,F m·Frame LO , LI,···, Ln-2, Ln-1 : information of each line of watermark TL1,...,TLn/2,TLn/2+l,...: the top line of the frame BL1,...,BLn/2, BLn/2+l,...: the bottom line of the frame