Implementation of Embedding and Extracting Invisible Watermarking

Abstract

The protection of ownership and prevention of unauthorized tampering with multimedia data (audio, video, image and text) have become important concerns. Image authentication verifies the originality of an image by detecting malicious manipulation; the ultimate goal to the watermark is retrieve the right owner information from the received data in a correct way. Digital watermarking is the process that embeds data called watermark into multimedia object such that watermark can be detected or extracted later only with appropriate decoding mechanism. In this work, an image is taken from color image (24 bits) type and from BMP file type and is converted into gray scale image (256 bits) and then converted into binary file by using one of filters (Sobel, Prewitt, Robert) to find edge detection of original file. Data storage process is performed in original image in edge points corresponding to the same place in a binary image. These edges are specified randomly based on location of the edge mod 3 and then specifying one of values (R, G, B) randomly to store data in it. As a result of this paper work invisible watermark is not noticeable to viewer and without any degrade the quality of the content. The product invisible watermark is robust against distortions processes and resistant to intentional tampering solely intended to remove the watermark. The embed information is repeated three times that take 2400 bits to keep on the quality image and even undistinguished in any image the watermark is embed.