GB2377107A

GB2377107A - Tamper detection method using selected pairs of pixels

Info

Publication number: GB2377107A
Application number: GB0115837A
Authority: GB
Inventors: Paola Marcella Hobson; Jonathan Stephen Hare
Original assignee: Motorola Inc
Current assignee: Motorola Solutions Inc
Priority date: 2001-06-28
Filing date: 2001-06-28
Publication date: 2002-12-31
Anticipated expiration: 2021-06-28
Also published as: GB2377107B; GB0115837D0

Abstract

A video communication unit (305) includes a video input (315) for receiving an un-watermarked video signal transmission having a number of video or image frames (110), the video input (315) being operably coupled to a processor (310). The processor (310) selects at least one pair of pixels (170, 180) from at least one video or image frame (110) and applies a rule-based mechanism to differentiate the selected at least one pair of pixels to apply a tamper evident means to the video or image signal transmission. A video communication system and method of detecting a tampered area and visually labelling the area are described. This enables fraudulent tampering of images and video to be detected, and the location of such tampering to be revealed to users of the material.

Description

Video/Image Communication With Watermarking Field of the Invention This invention relates to video transmission systems and related video encoding/decoding techniques. The invention is applicable to, but not limited to, a video compression system employing video watermarking where any tampering of a video image or portion of video image is to be detected.

Background of the Invention The ability to transmit real-time video and/or image data is a desirable characteristic of many current wireline and wireless communication systems. However, it is known that individual images/pictures, or a series of images say, in a transmitted video stream, may be subjected to'attacks', i. e. the images may have been tampered with. Therefore, a need exists to protect image or video transmissions from such undesirable tampering. One known technique employed to protect still/video images or documents is by the use of "watermarks".

In the context of the present invention, the terms'video' and'image'are used interchangeably, with the term'video' generally used to represent one or more still images.

Wolfgang R, Podilchuk C, Delp E"Perceptual watermarks for digital images and video", SPIE Conference on Security and Watermarking of Multimedia Content, January 1999, describes

some state of the art watermarking methods for use with video and images.

Protection of digital media (including image and video) has also become a key standardisation topic within the multimedia industry over the last year. Police users have formally stated that they do not envisage using digitally transmitted and processed images for evidential purposes without the existence of reliable tamper detection methods.

The European Broadcasting Union has issued a second call for systems that offer watermarking of multimedia transmissions for entertainment applications. In addition, the International Standards Organisation (ISO) has set up a working group known as MPEG-21, whose essential function is to investigate digital rights management including the authentication of multimedia data.

In image watermarking, a known binary pattern or signature is embedded into an image at the moment of image acquisition. Such watermarks are termed"robust", because they are designed to remain intact regardless of any postprocessing of the image such as filtering, cropping, etc.

While such watermarks do provide a useful degree of protection, they cannot at present be wholly relied on in a court of law. The purpose of these watermarking methods is such that they are not designed to possess the required degree of surety that an image has not been tampered with, in order for the image to be used as evidence.

Thus, there exists a need in the field of the present invention to provide a video communication unit and methods, based on a watermarking system, that can be used for testing a video sequence for evidence of tampering, wherein the abovementioned disadvantages may be alleviated.

Furthermore, there exists a need for a labelling method to highlight areas of a video sequence that are detected as having been tampered with.

Additionally, it would be beneficial to visually label tampered video sequences such that they are rendered unusable, or valueless, to the attacker.

Statement of Invention In accordance with a first aspect of the present invention there is provided a video communication unit, as claimed in claim 1.

In accordance with a second aspect of the present invention there is provided a video communication unit, as claimed in claim 7.

In accordance with a third aspect of the present invention there is provided a video communication unit, as claimed in claim 9.

In accordance with a fourth aspect of the present invention there is provided a video transmission system adapted to use one of the aforementioned video communication units, as claimed in claim 17.

In accordance with a fifth aspect of the present invention there is provided a mobile radio device, as claimed in claim 18.

In accordance with a sixth aspect of the present invention there is provided a method of watermarking a video signal transmission in a video transmission system, as claimed in claim 20.

In accordance with a seventh aspect of the present invention there is provided a method of detecting tampering of a watermarked digital image, as claimed in claim 27.

In accordance with an eighth aspect of the present invention there is provided a method of visually labelling a video sequence that contains an attacked watermark, as claimed in claim 29.

In accordance with a ninth aspect of the present invention there is provided a storage medium storing processorimplementable instructions for controlling a processor to carry out any of the aforementioned method steps of the sixth, seventh or eighth aspects of the present invention, as claimed in claim 35.

In accordance with a tenth aspect of the present invention there is provided a video communication unit adapted to perform any of the method steps of the aforementioned aspects, as claimed in claim 36.

In accordance with an eleventh aspect of the present invention there is provided a mobile radio device, as claimed in claim 37.

Brief Description of the Drawings Exemplary embodiments of the present invention will now be described, with reference to the accompanying drawings, in which: FIG. 1 shows a pixel-pair mechanism to determine whether an embedded watermark has been attacked, in accordance with the preferred embodiment of the invention.

FIG. 2 shows a flowchart of a decision process for determining whether tampering has occurred, in accordance with the preferred embodiment of the invention.

FIG. 3 shows a block diagram of a video communication system incorporating a communication unit embedding a watermark, and a communication unit detecting a watermark, in accordance with the preferred embodiment of the invention.

Description of Preferred Embodiments The inventive concepts described herein find particular application in the current MPEG Standards activities, where a standard watermarking system for video use is to be defined. The detection of tampering, and the ability to determine what type of tampering has taken place, are necessary steps in ensuring user confidence in the images and videos a user is viewing in a potentially hostile multimedia communication environment.

In summary, the preferred embodiment of this invention aims to pre-process video material such that detection of tampering can take place.

Most current image and video watermark methods focus on pre-processing video and images such that any embedded watermark can be recovered, regardless of tampering, for example in copyright applications. The method described below, however, provides for'fragile'watermarks that are destroyed when the images are altered.

The preferred embodiment of the invention utilises at least one pixel-pair in at least one video or image frame to provide a watermark, with the at least one pixel-pair being selected such that the values can be changed with little visual degradation on the image frame being communicated.

Furthermore, the method makes clear that content has been tampered with such that the person carrying out the

unauthorised processing realises that their actions are evident. As a consequence they cannot re-sell or distribute the video as an"original".

As an example, let us consider a person making"pirated" video files for unauthorised distribution via an internet web site. The preferred method, as described below, identifies where the video had been tampered with (e. g. in reformatting the video for web distribution) and applies a label to the tampered area, thereby rendering the video unsuitable for onward distribution.

The preferred embodiment of the present invention can be applied to video sequences consisting of at least two image frames. Furthermore, the preferred embodiment of the present invention can be applied to image formats including YCbCr (a standard representation of a colour image, as specified in ITU Rec 601), red/green/blue (RGB), or any single component (e. g. Y only) of an image format consisting of more than one component. Advantageously, the preferred embodiment of the present invention can also be applied in a restricted area or region of an image, or throughout the entire image.

The watermark arrangement of the preferred embodiment of the invention clearly shows when tampering has occurred, as tamper evident information is embedded into the video data stream.

Furthermore, the method allows the video player to locate the exact spatial and temporal position of the tampering as the video is being played.

Referring first to FIG. 1, a watermarking method 100 is shown, in accordance with the preferred embodiment of the invention. The watermarking method 100 includes a sequence of video/image frames such as frame 110. For clarity purposes only, one video/image frame is shown, Frame N 110, and its equivalent if tampered (tampered frame N) 120. In practice, many more video/image frames will be used in a video/image sequence.

The usual representation of image data is as a series of pixels, located as rows and columns of an image. Common image formats include representation of each pixel in a number of bits, from 6 to 12 bits per component of the image. As an example, each single component (R, G or B) of a colour image may have 8 bits per pixel. Alternatively, an infrared image may have 12 bits per pixel, represented as a luminance component.

In accordance with a preferred embodiment of the present invention, in the first frame, frame'N'110, of the sequence, at least one pair of pixels at locations 170,180 is selected in a place that will accept the values of the pixels being changed with little visual degradation. The method for selecting the at least one pixel-pair 170,180 must be able to withstand the effect of changing one or both of the pixel values. In this manner, the video player

can find the same locations in the watermarked video sequence.

Advantageously, the number of pairs of pixels selected can be used to control the sensitivity of the watermark embedding/detecting method. This is due to the fact that the more pairs of pixels 170,180 that are selected, the more sensitive the watermark is. However, conversely, the more pixel-pairs 170,180 that are selected, the greater the image degradation will be.

Referring to FIG. 1, the method of embedding a watermark 100 is explained now in greater detail. Let us denote the value of the first pixel of a pair by C 170 and the second pixel by D 180. The'value'here may be the luminance value, or may alternatively be the chrominance value. In fact, the value may be any numerical parameter associated with the pixel.

Each pair of pixels from the frame must be set so that the following rule applies:

This is then repeated in subsequent frames, for example frame'N+1', using preferably the same locations of pixelpairs 170,180 that were generated in the first frame'N' 110. The difference between the values of pixel-C 170 and pixel-D 180 also defines how sensitive the method is.

Thus, a larger difference results in a sequence in which small changes will be ignored.

As an example, select B locations in a frame of video, and designate the pixels at those locations as of pixel-type

'c'170. Also, select B corresponding locations in the same frame, spatially distant from those of pixel-type 170, designated pixel-type'0'180. The number of pixels of type C 170 and type D 180 must be equal. Also, the number of pixels of type C 170 must be less than half the total number of pixels in the frame'N'110.

I In the preferred embodiment of the invention the pixel locations of pixel-type C 170 are selected by a pseudo random number generator with a user defined private key, and the offsets to pixel locations of type D 180 are fixed or variable.

However, a skilled artisan would recognise that many other alternative location selection rules can be used, in order to achieve similar benefits to those obtained in the preferred embodiment. In addition, it is within the contemplation of the invention that any number of pixels can be used.

It is within the contemplation of the invention that alternative rules/pixel numbers/pixel arrangements etc. can be used to benefit from the inventive concepts described herein.

For each of the B pixels of type C 170, if the type'C' pixel value at each of the B location is greater than the

type'D'180 pixel values at the corresponding location, no changes are needed.

If the type C pixel value is less than the corresponding type'D'pixel value, then the preferred method is to either: (i) increase the value of the pixel-type C 170; and/or (ii) decrease the value of the pixel-type D 180.

(Note: It is not always possible to increase or decrease the value of a pixel if it takes it beyond the allowable limits for pixel values, i. e.'0'and'255'for an 8 bit per pixel per component image format).

The inventors of the present invention have recognised that the pixel locations generated from the first frame'N'110 may cease to be good locations for hiding the watermark in the current frame, and may start to cause visual artefacts.

In this regard, in accordance with the preferred embodiment of the present invention, the locations of at least one pair of pixels for watermarking may be regenerated using every, say, 3rd frame. This would increase computational complexity, but reduce visual artefact build-up.

When the frame is received at a video/image receiving unit a comparison is then made between pixel location'c'170 and pixel location'D'180. If the aforementioned rule [1]

still applies, with C > D, then it is assumed that the first frame 110 has not been tampered with.

However, if the aforementioned rule [1] does not apply, namely:

then it is assumed that the frame 120 has been tampered with.

It is within the contemplation of the invention that a video/image transmission system having any number of frames in a sequence of video/image frames, including any number of bits per pixel, would benefit from the inventive concepts described herein.

A skilled artisan would recognise that the robustness of this watermark can be manually tuned to suit the requirements of the video sequence.

The number of pixels per frame that are set according to the rule base mechanism may be varied. Similarly, the distribution of these pixels over the frame may be varied.

If a large number of pixels are set to satisfy the rule, and these pixels are spread well over the frame, then an attacker would have great difficulty in tampering with any part of the image. The chances of visual artifacts appearing in normal use of the image are however higher.

If however, few pixels are set in accordance with the rule and/or these pixels are all located in one region of the image, than an attacker is more likely to find a part of

the image which can be tampered with, without detection.

Setting fewer pixels in accordance with the rule will however reduce the likelihood of perceptible artifacts in the image.

Referring next to FIG. 2, a method 200 of detecting a watermark is shown, in accordance with the preferred embodiment of the invention.

In order to detect tampering of one or more frames, a

comparison is made between the value of pixel-'C'170 and the value of pixel'D'180, as shown in step 202. The comparison is made for each of the locations of the at least one pair of pixels to which the tamper evidence has been applied within a frame. The comparison is used to check that the embedding rule [1], i. e. that C > D still holds throughout the sequence, as shown in step 204.

Any pairs of pixels in any frames where the C > D relationship does not hold are assumed to have been tampered with, as in step 208. The approximate location of the tamper corresponds to the location of either C or D or both.

If the rule selected holds true, the video/image sequence is assumed to be tamper-free and the next frame/pixel values are selected for comparison, as shown in step 206.

If the comparison in step 204 does not yield a match, namely indicating tampering, in step 208, the frame may be

visually labelled as having been tampered with, as in step 210.

The visual labelling, in step 210, may take any form appropriate to make clear that tampering has occurred. The visual labelling step may include one or any combination of the following techniques: (i) replacing any or all of the tampered image pixels with a known value. The known value is preferably selected to be sufficiently different from the source image content such that the tampering is clearly visible, for example black, white, any saturated colour, any non-natural colour ; (ii) altering only the coloured appearance of a tampered pixel such that the underlying image content remains visible but the tampering is clearly marked; (iii) replacing only one component of a tampered pixel with a known value, for example 0 or 255, in an image format comprising more than one component; (iv) visually labelling (using one or more of (i) or (ii) or (iii) above) the complete image frame within which any of the pixels are detected as having been tampered with and/or; (v) visually labelling (using one or more of (i) or (ii) or (iii) above) the complete image frame, and all

subsequent images in the video sequence, within and following an image frame in which any of the pixels are detected as having been tampered with.

Once the tampered frame (s) have been visually labelled, the process takes the next pixel-pair for comparison, as in step 206.

Examples of'watermarking'communication units suitable for incorporating the aforementioned inventive concepts are described in co-filed patent applications GB9914384.4 and GB0031085.4, of the present applicant, whose contents are contained herein by reference.

However, FIG. 3 describes a preferred configuration of video/image communication units to implement the preferred embodiment of the present invention. Referring now to FIG.

3, a video communication system 300 is shown, in accordance with the preferred embodiment of the invention. The video communication system 300 includes a transmitting video/image communication unit 305 for embedding a watermark, and a receiving video/image communication unit 350 for detecting a watermark.

The transmitting video/image communication unit 305 includes a video/image input port 315 for receiving video/image signals. A video/image signal is passed to processor 310, which includes three watermark-embedding processes.

A first selection function/algorithm 320 selects the component (e. g. R or G or B or Y etc. ) in which to embed the tamper evidence.

A second selection function/algorithm 325 then selects the region, i. e. the frame (s) /pixel-pair (s), of the video/image in which to embed the tamper evidence. The second selection function/algorithm 320 may use a pseudo random generator 322 with a user defined key 324.

The tamper evidence is then applied in function/algorithm 330, in accordance with the method described with respect to FIG. 1. The video signal is then transmitted from transmitter 335 to the receiving video/image communication unit 350.

It is within the contemplation of the invention that alternative forms of communicating video or image data may be used, for example'transmission'may take the form of copying onto video tape, sending as an internet file, copying onto a floppy disk, etc.

The receiving video/image communication unit 350 includes a receiver 355 for receiving video/image signals. A received watermarked video/image signal is passed to processor 360, which includes three watermark processes.

A first function/algorithm 365 applies the tamper detection method in selecting the at least one pair of pixels, in accordance with the method described with respect to FIG.

1. The preferred first function/algorithm 365 uses a pseudo random generator 367 with a user defined key 368 to select the at least one pair of pixels for testing.

A second tamper detection function/algorithm 370 then detects whether tampering occurred, in accordance with the method described with respect to FIG. 1 and FIG. 2.

If tampering is detected, the video/image signal is passed to a third visually labelling function/algorithm 375, to label the tampered areas, prior to passing the tampered signal to a display 380.

A benefit of the aforementioned inventive concepts is that they can be readily implemented in existing video communication units. More generally, the set of algorithms used to effect the pixel-pair manipulation and processing may be implemented in a respective communication unit in any suitable manner. For example, new apparatus may be added to a conventional communication unit.

Alternatively existing parts of a conventional communication unit may be adapted, for example by reprogramming one or more processors therein. As such the required adaptation may be implemented in the form of processor-implementable instructions stored on a storage medium, such as a floppy disk, hard disk, programmable read only memory (PROM), random access memory (RAM) or any combination of these or other storage multimedia.

It will be understood that the video transmission and watermarking arrangements described above provide at least the following advantages: (i) means for submission of video evidence to a court of law for which it may be shown that the video has not been tampered with since initial acquisition and storage; (ii) means for purchasers of video and image material to authenticate the material such that it has not been altered since initial acquisition and storage; (iii) means for distributors of video and image material to verify that material passed to them for distribution is genuine and has not been tampered with since initial acquisition and storage; (iv) means for fraudulent tampering of images and video to be detected, and the location of such tampering to be revealed to users of the material; and (v) the robustness of the watermark can be manually tuned to suit the requirements of the video/image sequence.

The present invention allows a list of pixels to be provided to a receiving unit, this list indicating which pixels within the received image should satisfy the rule

based mechanism. This list needs only to be known to the receiver. It can be provided to the receiver in advance of any image transmission, and does not need to be sent along with the transmission of the image. Therefore there is less risk of an attacker knowing which pixels have had the rule based mechanism applied to them, which in turn further ensures the security of the invention.

Thus, a video communication unit has been provided wherein the abovementioned disadvantages associated with prior art arrangements have been alleviated. The video communication unit includes a video input for receiving an un-watermarked video signal transmission having a number of video or image frames. The video input is operably coupled to a processor. The processor selects at least one pair of pixels from at least one video or image frame and applies a rule-based mechanism to differentiate the selected at least one pair of pixels to apply a tamper evident means to the video or image signal transmission.

In addition, or in the alternative, a video communication unit includes a video receiver for receiving, from a transmitting video communication unit, a watermarked video signal transmission having a number of video or image frames. The receiver is operably coupled to a processor.

The processor compares at least one pair of pixels in at least one video or image frame to detect any tampering of the video or image frame.

In addition, or in the alternative, a video communication unit includes a processor that detects tampering of an area

of an image, and visually labels, upon detection of said tampering, said area to inform a user viewing the video of said tampering.

In such a case, it is not essential that the aforementioned method of tamper detection is used. The concept of visually labelling a tampered area, as hereinbefore described, can be coupled to any tamper detection method.

A mobile radio device, for example a mobile phone, a portable or mobile PMR radio, a personal digital assistant, a lap-top computer or a wirelessly networked PC may incorporate any of the aforementioned video communication units.

Furthermore, a method of watermarking a video signal transmission in a video transmission system has been described. The method includes the step of receiving an un-watermarked video or image signal transmission having a number of video or image frames. The method further includes the steps of selecting at least one pair of pixels from at least one video or image frame; and applying a rule based mechanism to differentiate the selected at least one pair of pixels to apply a tamper evident means to the video or image signal transmission.

In addition, or in the alternative, a method of detecting tampering of a watermarked digital image has been described. The method includes the steps of receiving a digitally watermarked image having a number of video or

image frames, wherein each video or image frame includes a number of pixel pairs, at least one of which had been modified. The method further includes the steps of comparing at least one pair of pixels in at least one video or image frame to detect any tampering of the video or image frame.

In addition, or in the alternative, a method of visually labelling a video sequence containing an attacked watermark has been described. The method includes the steps of detecting tampering of an area of an image and visually labelling the tampered area to inform a user viewing the video of the tampering.

Again, in such a case, it is not essential that the aforementioned method of tamper detection is used. The concept of visually labelling a tampered area, as hereinbefore described, can be coupled to any tamper detection method.

Further, a storage medium, storing processor-implementable instructions for controlling one or more processors to carry out the above-mentioned method steps, has been described.

In such a manner, the disadvantages associated with prior art arrangements have been alleviated.

Claims

Claims 1. A video communication unit (305) comprising a video input (315) for receiving an un-watermarked video signal transmission having a number of video or image frames (110), wherein each video or image frame includes a number of pixels, the video input (315) being operably coupled to a processor (310), the video communication unit (305) characterised by said processor (310) selecting at least one pair of pixels (170,180) from at least one video or image frame (110) and applying a rule based mechanism to differentiate the selected at least one pair of pixels to apply a tamper evident means to the video or image signal transmission.
2. The video communication unit (305) according to claim 1, wherein at least two consecutive frames are arranged such that they include the same watermarked pixelpair locations (170,180).
3. The video communication unit (305) according to claim 1 or claim 2, wherein the at least one pair of pixel values are selected based on their suitability for modification without causing visual degradation.
4. The video communication unit according to any preceding claim, wherein the at least one pair of pixels are selected by a pseudo random generator (324) with a user defined key (322).

<Desc/Clms Page number 23>
5. The video communication unit according to any preceding claim, wherein the number of pairs of pixels selected on any particular frame is variable dependent upon the sensitivity of the desired watermark against tolerable image degradation.
6. The video communication unit (305) according to any preceding claim, wherein a first selected pixel of a pair is denoted by a first value and the processor, in response to the rule based mechanism, adapts a value of a second pixel of the pair to be less than the value of the first pixel of the pair or adapts a value of a first pixel of the pair to be more than the value of the second pixel of the pair.

<Desc/Clms Page number 24>
7. A video communication unit (350), in particular a video communication unit according to any preceding claim, comprising a video receiver (355) for receiving, from a transmitting video communication unit (305), a watermarked video signal transmission having a number of video or image frames, the receiver being operably coupled to a processor (360), the video communication unit (350) characterised by said processor (360) comparing (202) a value of at least one pair of pixel locations in at least one video or image frame to detect any tampering of the video or image frame.
8. The video communication unit (350) according to claim 7, further characterised by said processor (360) determining (204) whether a frame has been tampered with by comparing the at least two pixel values, wherein: if said comparison shows that a rule based mechanism applied to embed the watermark still applies, it is assumed that no tampering of the video or image frame occurred (110); or if said comparison shows that the rule based mechanism applied to embed the watermark does not apply, it is assumed that tampering of the video or image frame occurred (120).
9. A video communication unit (350), in particular according to the video communication unit of claim 7 or claim 8, the video communication unit comprising a processor (360) that detects tampering of an area of an image, the video communication unit characterised by said processor visually labelling (375), upon detection of said

<Desc/Clms Page number 25>

tampering, said area to inform a user viewing the video of said tampering.
10. The video communication unit (350) according to claim 9, wherein said visual labelling (375) includes replacing any or all of the tampered image with a known value such that the tampering is visible, for example black, white, any saturated colour, and/or any non-natural colour.
11. The video communication unit (350) according to claim 9, wherein said visual labelling (375) includes altering only a coloured appearance of a tampered pixel such that the underlying image content remains visible but the tampering is marked.
12. The video communication unit (350) according to claim 9, wherein said visual labelling (375) includes replacing one component of a tampered pixel with a known value in an image format comprising more than one component.
13. The video communication unit (350) according to any of claims 9 to 12, wherein a complete image frame is visually labelled (375), when any of the pixels within said frame are detected as having been tampered with.
14. The video communication unit (350) according to claim 13, wherein the complete image frame and all subsequent images in the video sequence within and

<Desc/Clms Page number 26>

following an image frame in which any of the pixels are detected as having been tampered with are visually labelled.
15. The video communication unit (305,350) according to any preceding claim, wherein the watermark is applied to at least one of the following image formats: YCbCr, RGB, or any single component of an image format.
16. The video communication unit (305, 350) according to any preceding claim wherein the watermark is applied in a restricted area or region of an image, or throughout the entire image.
17. A video transmission system (300) adapted to use the video communication unit of any of the preceding claims.
18. A mobile radio device comprising a video communication unit (305,350) in accordance with any of claims 1 to 16.
19. The mobile radio device of claim 18, wherein the mobile radio device is a mobile phone, a portable or mobile PMR radio, a personal digital assistant, a lap-top computer or a wirelessly networked PC.

<Desc/Clms Page number 27>
20. A method of watermarking (100) a video signal transmission in a video transmission system, the method comprising the step of: receiving (315) an un-watermarked video or image signal transmission having a number of video or image frames, wherein each video or image frame includes a number of pixels; the method characterised by the steps of: selecting at least one pair of pixel locations (170,180) from at least one video or image frame (110); and applying a rule based mechanism to differentiate the selected at least one pair of pixels to apply a tamper evident means to the video or image signal transmission.
21. The method of watermarking (100) a video signal transmission according to claim 20, wherein the step of selecting includes the step of: selecting at least one pair of pixels from at least one frame of the video or image frame (110).
22. The method of watermarking (100) a video signal transmission according to claim 21, wherein the step of selecting is made in at least two consecutive frames such that the at least two consecutive frames include the same pixel-pair locations (170,180).

<Desc/Clms Page number 28>
23. The method of watermarking (100) a video signal transmission according to any of claims 20 to 22, wherein the step of selecting includes selecting at least one pair of pixel values based on their suitability for modification without causing visual degradation.
24. The method of watermarking (100) a video signal transmission according to any of claims 20 to 23, wherein the step of selecting at least one pair of pixels is made using a pseudo random generator (324) with a user defined key (322).
25. The method of watermarking (100) a video signal transmission according to any of claims 20 to 24, wherein the step of selecting includes the step of: selecting a variable number of pixel-pairs to be used for the watermark in at least one frame dependent upon at least one of: the sensitivity of the desired watermark or tolerable image degradation.
26. The method of watermarking (100) a video signal transmission according to any of claims 20 to 25, wherein the step of selecting includes the steps of: selecting a first pixel; denoting the first pixel by a first value; and applying a rule based mechanism to adapt a value of a second pixel of the pair to be less than the value of the first pixel of the pair or adapt a value of a first pixel of the pair to be more than the value of the second pixel of the pair.

<Desc/Clms Page number 29>
27. A method of detecting tampering (200) of a watermarked digital image, the method comprising the steps of: receiving (355) a digitally watermarked image having a number of video or image frames, at least one of which is watermarked; the method characterised by the steps of: extracting (365) determined pixel locations from at least one video or image frame; and comparing (202) at least one pair of pixels in at least one video or image frame to detect any tampering of the video or image frame.
28. The method of detecting tampering (200) of a watermarked digital image according to claim 27, wherein the step of detecting includes the step of: determining whether a frame has been tampered with by comparing the at least two pixel values, wherein: if said comparison shows that a rule based mechanism applied to embed the watermark still applies, it is assumed that no tampering of the video or image frame occurred (110); or if said comparison shows that the rule based mechanism applied to embed the watermark does not apply, it is assumed that tampering of the video or image frame occurred (120).

<Desc/Clms Page number 30>
29. A method of visually labelling (375) a video sequence containing an attacked watermark, in particular according to the method of detecting tampering of a watermarked digital image of claims 27 or claim 28, the method comprising the step of: detecting tampering (370) of an area of an image; and characterised by the step of: visually labelling (375) said area to inform a user viewing the video of said tampering.
30. A method of visually labelling (375) a video sequence according to claim 29, the method further characterised by the step of: altering a coloured appearance of a tampered pixel to inform a user viewing the video/image sequence of said tampering.
31. The method of visually labelling (375) a video sequence according to claim 29, wherein said visual labelling step includes the step of: replacing any or all of a tampered image with a known value such that the tampering is visible, for example black, white, any saturated colour, and/or any non-natural colour.

<Desc/Clms Page number 31>
32. The method of visually labelling (375) a video sequence according to claim 29, wherein said visual labelling step includes the step of: altering only a coloured appearance of a tampered pixel such that the underlying image content remains visible but the tampering is marked.
33. The method of visually labelling (375) a video sequence according to claim 29, wherein said visual labelling step includes the step of: replacing one component of a tampered pixel with a known value in an image format comprising more than one component.
34. The method of visually labelling (375) a video/image sequence according to any of claims 29 to 33, wherein said visual labelling step includes the step of: visually labelling a complete image frame within which any of the pixels are detected as having been tampered with.
35. A storage medium storing processor-implementable instructions for controlling one or more processors (310, 360) to carry out the method of any of claims 20 to 34.
36. A video communication unit (305,350) adapted to perform any of the method steps of any of preceding claims 20 to 34.

<Desc/Clms Page number 32>
37. A mobile radio device comprising a video communication unit in accordance with claim 36.
38. The mobile radio device of claim 37, wherein the mobile radio device is a mobile phone, a portable or mobile PMR radio, a personal digital assistant, a lap-top computer or a wirelessly networked PC.
39. A video communication unit or method of embedding or detecting a watermark in a video sequence substantially as hereinbefore described with reference to, and/or as illustrated by, the accompanying drawings.