Data processing apparatus and method

A method for an image processing apparatus of detecting at least one of a set of code words which has been embedded in a water marked version of an image, the method comprising:

• registering, with the image processing apparatus, the water marked version of the image and an original version of the image with respect to one another;
• recovering, with the image processing apparatus, an estimate of the at least one of the set of code words from the water marked version of the image by comparing the water marked version of the image with the original version of the image;
• and
• detecting the at least one of the set of code words with the image processing apparatus by correlating each of the set of code words with the at least one of the set of code words, wherein
• the registering the water marked version of the image and the original version of the image comprises,
• dividing, with the image processing apparatus, a frame of the water marked version of the image into a plurality of image blocks,
• estimating, with the image processing apparatus, for each of the plurality of image blocks of the water marked version of the image, an observed probability of possible distortion vectors representing shifts of the image block with respect to positions of the image block in the original version of the image,
• generating, with the image processing apparatus, for each of the plurality of image blocks, an estimate of a forward probability distribution of the possible distortion vectors within the image block recursively from the observed probability of possible distortion vectors for the image block and forward probability distribution estimates determined from previous image blocks for a predetermined path through the plurality of image blocks,
• generating, with the image processing apparatus, for each of the plurality of image blocks, an estimate of a backward probability distribution of the possible distortion vectors within the image block recursively from the observed probability of possible distortion vectors for the image block and backward probability distribution estimates determined from subsequent image blocks for a predetermined pattern,
• combining for each of the plurality of image blocks the estimate of the forward probability distribution, the estimate of the backward probability distribution, and the observed probability of possible distortion vectors with the image processing apparatus to form a most likely distortion vector estimate,
• forming, with the image processing apparatus, a map of distortion vectors from the most likely distortion vector estimate for each of the plurality of image blocks, and
• applying, with the image processing apparatus, a reverse of the most likely distortion vector for each of the plurality of image blocks by re-sampling the water marked version of the image to register the water marked version of the image with the original version of the image to detect the at least one of the set of code words.

A computer readable medium having recorded thereon a computer program providing computer executable instructions, which when loaded on to a data processor, cause the data processor to perform a method comprising:

• registering a water marked version of an image and an original version of the image with respect to one another;
• recovering an estimate of at least one of a set of code words embedded in the water marked version of the image by comparing the water marked version of the image with the original version of the image;
• and
• detecting the at least one of the set of code words by correlating each of the set of code words with the at least one of the set of code words, wherein
• the registering the water marked version of the image and the original version of the image includes,
• dividing a frame of the water marked version of the image into a plurality of image blocks,
• estimating, with the data processor, for each of the plurality of image blocks of the water marked version of the image, an observed probability of possible distortion vectors representing shifts of the image block with respect to positions of the image block in the original version of the image,
• generating, with the data processor, for each of the plurality of image blocks, an estimate of a forward probability distribution of the possible distortion vectors within the image block recursively from the observed probability of possible distortion vectors for the image block and forward probability distribution estimates determined from previous image blocks for a predetermined path through the plurality of image blocks,
• generating, with the data processor, for each of the plurality of image blocks, an estimate of a backward probability distribution of the possible distortion vectors within the image block recursively from the observed probability of possible distortion vectors for the image block and backward probability distribution estimates determined from subsequent image blocks for a predetermined pattern,
• combining for each of the plurality of image blocks the estimate of the forward probability distribution, the estimate of the backward probability distribution, and the observed probability of possible distortion vectors, with the data processor, to form a most likely distortion vector estimate,
• forming a map of distortion vectors from the most likely distortion vector estimate for each of the plurality of image blocks, and
• applying a reverse of the most likely distortion vector for each of the plurality of image blocks by re-sampling the water marked version of the image to register the water marked version of the image with the original version of the image to detect the at least one of the set of code words.