علياء محسن ضايع عجينة

The growing use of digital images in a wide range of applications, and growing the availability of many editing photo software, cause to emerge a challenge to discover the images tampering. In this paper, we proposed a method to detect the most important type of forgery image (copy and move). We suggested many steps to classify the image as forgery or non-forgery image, started with preprocessing (included, convert image to gray image, de-noising, and image resize). Then, the image will be divided into several overlapping blocks. For each block, feature extracted (used it as a matching feature) by using the singular value decomposition (SVD) transformation. According to these features, the pixels were collected in many main groups, and then these groups clustered to many subgroups. The weight for each main group can be determined by comparing the subgroups with each other according to suggested conditions. The number of subgroups and weights are used to classify images to forgery or non-forgery images. The accuracy of detection and classified the forgery images were up to 97%. The suggested method is robust for tampered object rotation, scaling, and change of illumination.

Information security has become the topic of most interest nowadays. Because of significant technological development, transferring information from one person to another has become very easy. This information may be exposed to penetration by unauthorized persons, so the security interest in encryption and steganography has increased. In this paper presents a novel approach by integrating the methodologies of steganography and encryption, resulting in an algorithm that exhibits enhanced resilience against external attacks. The key is first initialized using Grigorchuk's group, which gives an unstructured permutation of data bits. Followed by the stage of encrypting the audio file using RC4, which relies on random generation of the final key in which it is encrypted. In order to increase security and protect data from hacking, the encrypted audio file is hidden in video frames using a non-sequential method based on LSB. The proposed method proved effective results in data preservation as it provided the Peak Signal-to-Noise Ratio scale value of 88, which makes high imputation capacity.

A biometric system allows an individual to be automatic identification using a distinguishing or single feature possessed by the person. The biometric system of identification available which is regarded as the most accurate and reliable known is iris recognition. In this paper, we discuss the strategies used to construct an Iris Recognition System, as well as an analysis of our findings. To locate the limits of the iris in the digital image of the human eye, we used an integration procedure that incorporated both a Polar Spline RANSAC and a Total Variation Model. Predictable patterns of an individual's iris are retrieved as a feature vector using Daugman's rubber sheet model and the Gabor filter. The quantified values are then compared using the Hamming Distance operator to see whether the two irises are really the same. Experiments demonstrated that the accuracy of the recommended strategy for photographs acquired in uncooperative situations is either superior to or equivalent to other ways provided in the literature.

Human skin detection is a process of classification of the image pixels or regions as skin or non-skin. Skin detection has many applications such as face tracking, skin diseases, nudity recognition, hand gesture, video surveillance, web content filtering, and pornographic content filtering. Skin detection is a challenging problem due to skin color variation, the human race, aging, gender, makeup, complex background, etc. This paper suggests detecting the skin region in the image and finding the location of the skin based on a convolution neural network. In this proposal, the proposed CNN will be modified by adding two layers before the first layer of CNN and after the last layer of CNN. The main purpose of these layers is to prepare the input image by using a sliding window which is input an indexed small part of the image into the CNN. The network classifies each part as skin or non-skin and then sent the result into the second suggested layer. After processing all the image pixels, the non-skin blocks are mapped to the original image as black regions. The final image contains the skin regions with black color in the background. The contribution of the proposed method is the ability to detect the skin from any part of the human body, unlike the previous works which focus on one part of the body, the input image process as blocks instead of processing the entire images as in the previous works and then in the output reconstruct the original image, this method works well with most of the challenges that face the detection of skin, and finally, the designed network facilitates the localization and segmented of the skin region almost accurately, while the previous networks focus on classification the image as skin or non-skin. The accuracy of the detection of skin when testing with images different from the training images was 95.4%.