| المستخلص: |
After the great progress in editing software for digital images and facilitated by a broad and diversified and with the cost of low-lying and high efficiency and ease of working out, making the subject of credibility and integrity of digital images and verify the contents is it original or has been manipulated is very important and necessary, and it has become important to ensure the safety of digital and credibility of the content of objects and image information content, using programs based on image analysis or add and hide what confirms the correctness of this image through the creation of these programs and technology that helps detect fraud and manipulation in the digital images. A proposed system is designed to detect the manipulation of the image in any part of it down to the level of accuracy in high discovered through the use of has hand coding techniques for each part of the image where the image is divided into number of part (depend on the size of the image) then hide the hash within the image parts. In this proposed system has been used techniques the (SHA-512 and the encoding RLE) for generating the digital signature of each part of the image. The system consists of two phases of the has hand encoding, are in the first Phase hash tone account previous methods for each section of the image after making process (Haarwave let) by and enter the code and the number of partition with the calculated hash to the second phase of the hash and the result will be encrypt and embedded And hidden in the color layers within the image. By studying the results of the Proposed system and found that the rate of discovery of any kind of manipulations used in the screening process, which included any type from copy paste)and resizing to reduce or increase the size image or object in the image ( stretching for the image and Rotating for all the image or just for one object in the image in any degree and removing or adding object from the image and also the enhancement to the image etc. The speed of concealment and discovery for the proposed method depend on the number of the image parts, while the fastest results arrived infractions of a second for the color image.
|
