Omar Akif

Dr. Omar Akif


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 November 2022 in Sensors
Verified
Journal Article
Open access
PDF file
RARasha Subhi Ali
Rasha Subhi Ali
OAOmar Akif
Omar Akif
SJSameeh A. Jassim
Sameeh A. Jassim
Alaa kadhim FarhanAlaa kadhim Farhan
Alaa kadhim Farhan
El-Sayed M. El-KenawyEl-Sayed M. El-Kenawy
El-Sayed M. El-Kenawy
Abdelaziz A. AbdelhamidAbdelaziz A. Abdelhamid
Abdelaziz A. Abdelhamid
Background and Aim: due to the rapid growth of data communication and multimedia system applications, security becomes a critical issue in the communication and storage of images. This study aims to improve encryption and decryption for various types of images by decreasing time consumption and strengthening security. Methodology: An algorithm is proposed for encrypting images based on the Carlisle Adams and Stafford Tavares CAST block cipher algorithm with 3D and 2D logistic maps. A chaotic function that increases the randomness in the encrypted data and images, thereby breaking the relation sequence through the encryption procedure, is introduced. The time is decreased by using three secure and private S-Boxes rather than using six S-Boxes, as in the traditional method. Moreover, the CAST encryption algorithm was modified to be used on the private keys and substitution stage (S-Boxes), with the keys and S-Boxes of the encryption algorithm being generated according to the 2D and 3D chaotic map functions. The proposed system passed all evaluation criteria, including (MSE, PSNR, EQ, MD, SC, NC, AD, SNR, SIM, MAE, Time, CC, Entropy, and histograms). Results: Moreover, the results also illustrate that the created S-Boxes passed all evaluation criteria; compared with the results of the traditional method that was used in creating S-Box, the proposed method achieved better results than other methods used in the other works. The proposed solution improves the entropy which is between (7.991–7.999), reduces the processing time which is between (0.5–11 s/Images), and improves NCPR, which is between (0.991–1). Conclusions: The proposed solution focuses on reducing the total processing time for encryption and decryption and improving transmission security. Finally, this solution provides a fast security system for surgical telepresence with secure real-time communication. The complexity of this work needs to know the S-Box creation method used, the chaotic method, the values of the chaotic parameters, and which of these methods was used in the encryption process.

1 August 2024 in Electronics
Verified
Journal Article
Open access
PDF file
ZAZena Mohammad Al-badry
Zena Mohammad Al-badry
ASAhmed T. Sadiq
Ahmed T. Sadiq
OAOmar Akif
Omar Akif
Alaa kadhim FarhanAlaa kadhim Farhan
Alaa kadhim Farhan
As technology advances and develops, the need for strong and simple authentication mechanisms that can help protect data intensifies. The contemporary approach to giving access control is through graphical passwords comprising images, patterns, or graphical items. The objective of this review was to determine the documented security risks that are related to the use of graphical passwords, together with the measures that have been taken to prevent them. The review was intended to present an extensive literature review of the subject matter on graphical password protection and to point toward potential future research directions. Many attacks, such as shoulder surfing attacks, SQL injection attacks, and spyware attacks, can easily exploit the graphical password scheme, which is one of the most widely used. To counter these security threats, several measures have been suggested, but none of the security attacks can be completely overcome. Each of the proposed measures has its pros and cons. This study begins by elucidating some of the graphical password schemes studied between 2012 and 2023, delving into potential threats and defense mechanisms associated with these schemes. Following a thorough identification and selection process, five of the reviewed papers explain the threat of shoulder surfing and spyware attacks on graphical password schemes, while two explain the threat of brute force attacks. One paper focuses on dictionary attacks, while four other papers address social engineering, SQL injection attacks, and guessing attacks as potential threats to graphical password schemes. In addition, the papers recognize other forms of attacks, such as video recording attacks, filtering attacks, reverse engineering attacks, multiple observation attacks, key/mouse logger attacks, insider attacks, computer vision attacks, image gallery attacks, sonar attacks, reply attacks, data interception attacks, and histogram manipulation attacks. These attacks are examined in three, three, eight, one, four, one, one, one, one, one, one, and one papers, respectively. Moreover, out of all such countermeasures, most of them are based on three categories—randomization, obfuscation, and password space complexity—which are the most commonly employed strategies for improving graphical password schemes.

24 May 2025 in Iraqi Journal for Computer Science and Mathematics
Verified
Journal Article
Open access
PDF file
OAOmar Akif
Omar Akif
SASura Mazin Ali
Sura Mazin Ali
ASAnn F. Sabih
Ann F. Sabih
ASAhmed T. Sadiq
Ahmed T. Sadiq
SSS. K. Subramaniam
S. K. Subramaniam
An intrusion detection system (IDS) is key to having a comprehensive cybersecurity solution against any attack, and artificial intelligence techniques have been combined with all the features of the IoT to improve security. In response to this, in this research, an IDS technique driven by a modified random forest algorithm has been formulated to improve the system for IoT. To this end, the target is made as one-hot encoding, bootstrapping with less redundancy, adding a hybrid features selection method into the random forest algorithm, and modifying the ranking stage in the random forest algorithm. Furthermore, three datasets have been used in this research, IoTID20, UNSW-NB15, and IoT-23. The results are compared with the three datasets mentioned above and it emerges that the accuracy of the proposed system is 96.2%, which is better than the other methods in the IoTID20 Dataset, while the accuracy with the second dataset UNSW-NB15 yielded 98.85%. Lastly, using the third dataset, IoT-23, the suggested technique achieved 99.93%.

1 January 2025
Verified
Journal Article
Closed access
PDF file
ZZZena Zena
Zena Zena
ASAhmed T. Sadiq
Ahmed T. Sadiq
OAOmar Akif
Omar Akif
El-Sayed M. El-KenawyEl-Sayed M. El-Kenawy
El-Sayed M. El-Kenawy
ZAZena Mohammad Al-badry
Zena Mohammad Al-badry
With the continuous progress of image retrieval technology, the speed of searching for the required image from a large amount of image data has become an important issue. Convolutional neural networks (CNNs) have been used in image retrieval. However, many image retrieval systems based on CNNs have poor ability to express image features. Content-based Image Retrieval (CBIR) is a method of finding desired images from image databases. However, CBIR suffers from lower accuracy in retrieving images from large-scale image databases. In this paper, the proposed system is an improvement of the convolutional neural network for greater accuracy and a machine learning tool that can be used for automatic image retrieval. It includes two phases; the first phase (offline processing) consist of two stages; stage1 for CNN model classification while stage 2 for extracts high-level features directly from CNN by a flattening layer, which will be stored into a vector. In the second phase (online processing), the retrieval depends on query by image (QBI) from the system, which relies on the online CNN model stage to extract the features of the transmitted image. Afterward, an evaluation is conducted between the extracted features and the features that were previously stored by employing the Hamming distance to return all similar images. Last, it retrieves all the images and sends them to the system. Classification for images was achieved with 97.94% deep learning results, while for retrieved images, the deep learning was 98.94%. For this paper, work done on COREL image dataset. The images in the dataset used for training are more difficult than image classification due to the need for more computational resources. In the experimental part, training images using CNN achieved high accuracy, proving that the model has high accuracy in image retrieval.

1 January 2024 in Journal of Soft Computing and Computer Applications
Verified
Journal Article
Open access
PDF file
ZAZena Mohammad Al-badry
Zena Mohammad Al-badry
ASAhmed T. Sadiq
Ahmed T. Sadiq
OAOmar Akif
Omar Akif
MEMarwa M. Eid
Marwa M. Eid
In recent years, with the rapid development of the current classification system in digital content identification, automatic classification of images has become the most challenging task in the field of computer vision. As can be seen, vision is quite challenging for a system to automatically understand and analyze images, as compared to the vision of humans. Some research papers have been done to address the issue in the low-level current classification system, but the output was restricted only to basic image features. However, similarly, the approaches fail to accurately classify images. For the results expected in this field, such as computer vision, this study proposes a deep learning approach that utilizes a deep learning algorithm. In this research, a proposed model based on a Convolutional Neural Network (CNN) which is a machine learning tool that can be used for the automatic classification of images. The model is concerned with the classification of images, and for this, it employs the COREL Image dataset (Corel Gallery Image Dataset) as a reference. The images in the dataset used for training are harder than the classification of the images since they need more computational resources. In the experimental part, training the images using the CNN network achieved 98.52% accuracy, proving that the model has high accuracy in the classification of images.