The current research enhances the performance of an already proposed image protection system presented earlier. The former system relies on an association technique instead of embedding owner signature inside an image like traditional watermarking methods. Thus, it retains secured image characteristics against any type of tampering two additional units are involved at the second phase of the former system operation. The main functions of those two units are to resize and reform the image as a preprocessing before submitting it to common processing of signature generation. Neural network structures are used to build those two units. A method of data compression is used to save time and generalization feature of neural network. Samples of pixels with their encoded coordinates are used to construct a data base used for the training purpose where the relative coordinate is used instead of the absolute address of the pixels and then encoded to determine a useful data suitable for neural network processing. Major conclusion remarks are discussed to highlight the most significant recommendations needed for future work.

Watermarking Copyright Digital Signature.