- Biometric Identification and Security
- Chaos-based Image/Signal Encryption
- Network Security and Intrusion Detection
- Advanced Steganography and Watermarking Techniques
- Cloud Data Security Solutions
- Image and Video Stabilization
- Digital Media Forensic Detection
- Traditional Chinese Medicine Studies
- Image Retrieval and Classification Techniques
- Face and Expression Recognition
- Vehicle License Plate Recognition
- Face recognition and analysis
- Advanced Malware Detection Techniques
- Infrared Thermography in Medicine
- IoT and Edge/Fog Computing
- Cryptography and Residue Arithmetic
- Energy Efficient Wireless Sensor Networks
- Mobile Ad Hoc Networks
- Cloud Computing and Resource Management
- Flavonoids in Medical Research
- Radiomics and Machine Learning in Medical Imaging
- Cryptography and Data Security
- Security in Wireless Sensor Networks
- Privacy-Preserving Technologies in Data
- IoT-based Smart Home Systems
Mustansiriyah University
2019-2023
University of Baghdad
2018
Cloud <span>computing provides advantages, like flexibly of space, security, cost optimization, accessibility from any remote location. Because this factor cloud computing is emerging as in primary data storage for individuals well organisations. At the same time, privacy preservation an also a significant aspect computing. In regrades to preservation, association rule mining was proposed by previous researches protect users. However, algorithm involves creation fake transaction and...
Cloud computing complexity is growing rapidly with the advancements that it witnessing. It has created a requirement to simplify process of configuring cloud and re-configuring when required, also involves tasks like auto scaling infrastructure, elastic maintaining health servers. The proposed method introduces smart management using knowledge base, which models resources cloud; handles service level agreement its evaluations. base supports representational state transfer (REST/RESTful)...
The objective of feature extraction is to decrease the original data set by measuring definite features, or properties, which recognizes one input pattern from another.The main idea proposed system depends on where uses two phases first phase discrete wavelet transform and second seven moment’s invariants. In this paper apply level medical image; obtain bands texture features are extracted coefficients then moments invariant for each band 49 image. was implemented a real human MRI dataset,...
In this paper an Iris image is encrypted based on QR (quick response) code and chaotic map. The main idea of the proposed system generating a depending input text then extract features from by using convolution, these are used for key generation. After that permuted iris generated key, after resulting will be encrypts 2D logistic randomness tested measures NIST, quality images in method security analysis tests such as PSNR, UACI, NPCR, histogram, correlation entropy. shows secure encryption.
Today, in this new era of the internet, Information Security is becoming biggest challenge for world due to rapid growth internet users day by day.Unauthorized access secret data can have serious repercussions like a financial loss.One best techniques secure communication Steganography-or covert writing.It an art hiding very existence communicating message itself.This research presents state dual stenographic technique.Dual Steganography combines two security mechanisms, steganography and...
The cloud computing providing to users and companies many services facilities such as less cost, availability risk management. of are delivered in pay per usage way on-demand it can store huge amount data. But the security concern is main reason that makes afraid using their data on servers. In this paper, we try review describe encryption methods used protect which stored make a comparison among these methods.
In order to protection computers, programs, networks and data from intrusions unauthorised access (UA), alteration, or demolition, a set of technologies procedures is known as cybersecurity. A significant concern the identification prevention network intrusion. Methods like machine learning & deep identify through estimating risk utilizing training data. Through years, number techniques have been introduced, it has demonstrated that these are more accurate than other intrusion detection...
Attacks against cyber security are increasing in frequency and complexity over the years.Due to level of sophistication that increasing, defensive strategies need be developed further continuously innovated.Traditional intrusion detection deep packet inspection continue utilized suggested, but they adequate for addressing requirements changing threats.As processing power increases cost doing so drops, machine learning is seen as a complementary method or additional defense malware, botnets,...
Cloud computing (CC) has become more and popular in environments that use distributed computing. Cloud-based processing data storage is becoming a global trend. Perhaps might put up the effect Software as Service (SaaS) on numerous commercial applications daily lives disruptive technology. an extension of Internet-based computing, where devices can access shared software, resources, required. Machine learning (ML) tool areaof cybersecurity(CS)through cloud environment recent years. Examples...
Security concerns are multiplying as a result of the rapid advancement computer and communications technology. Cybersecurity is evolving into different types techniques to reduce these concerns. Amongst anomaly detection (AD). Anomaly identification (AI) major problem in many academic fields practical applications. It has been presented that Convolution Neural Networks (CNNs) be used identify anomalies from type; however, there currently no guide over which model apply specific instance. In...
The objective of feature extraction is the transformation input data into a set features. Features are distinctive properties patterns that help in differentiating between categories patterns. main idea proposed system depends on where uses wavelet transforms. In this paper apply colour information (RGB space) specific form two-level discrete transforms (DWT) Retain image; obtain seven bands texture features extracted from coefficients. paper, algorithm extracting retaining images person to...
Essentially use iris technique to human identification and recognition. Due enormous growth through the recent years for recognition techniques numerous research has emerged in areas like; image compression, segmentation, quality assessment, Image Acquisition (IA), restoration, feature extraction (FE), normalization, noise reduction code matching (NRCM), evaluation, searching large database, applications, performance under changing multibiometrics condition. Where many were suggested (IR)...
Due to the tremendous developments in communication networks especially Internet network, which used exchange many types of data by people. Data security has become a major concern. Therefore, there is an increasing interest use encryption and decryption methods. Many methods have been developed maintain security, one these that commonly recent are chaotic systems. Where were suggested encrypt images using map because their features such as sensitivity initial conditions unpredictability...
Because of the advantages resource sharing in an efficient and effective way, practical data storage mechanisms, resources delivered on demand whenever they are needed adequate suitable environment, cloud computing plays indisputable role today's software.The available is going to be frequently shared via various customers under different categorizations like public cloud, private hybrid so on, as stated National Institute Standards Technology (NIST).Generally, access control can defined a...
Physiological biometrics is one of the attractive fields for researchers due to its unique and stable features.One these physical "Palm Print".This new approach used in personal recognition because powerful information that can be extracted from palm print.The characteristics are keeping rules palmprint feature extraction very important.In spite huge work done this approach, results recorded about print still uncompleted techniques continually modified.In paper, we present a detailed...
Edge detection is a technique for detecting the presence and location of sharp changes in an image's intensity.Edges help segmentation object recognition by defining boundaries between sections image.Edge minimizes quantity data image filters out unnecessary information while keeping crucial structural qualities.The general method edge to study single pixel area, use variation neighboring first-order or second-order detect edge.This paper overview different techniques like differential...
Hand print recognition system received great interest in the recent few years such as human-computer interaction, computer vision, and graphics. In this paper, proposed for human handprint based on FP-growth algorithm, consists of three-stage. The first stage detection algorithm using HSV color space, canny contrast enhancement grayscale. separate skin area in-handprint image through space converting RGB to well conducting specific rules determining area. And then applies hand segmentation...
This paper, proposes the pre-processing methods use a small neighborhood of pixel in input image to get new brightness value output image.Such operations are also called filtration. this we start with using medical case for psoriasis after change it gray state implemented under transform domain (i.e frequency), wavelet then three filters sharpening, Sobel, and Laplace filter.after make proposed by computing PSNR each show effect it.Then extract features through an apply set measures (Energy,...
Due to the network and multimedia application developments, information security become important since can be attacked over transmission channels, combination of cryptography chaotic an field security, where many encryption algorithms are based on mapping due inherent features Image like high redundancy bulk data capacity.In this paper, three maps used achieve a good diffused by setting initial conditions generate shuffling sequence randomly 2D Zaslevskii map employ them in cat for image...
In this paper, proposed detection algorithm for human ear print images, the consist of three-stage.the first stage The using HSV color space, canny and contrast enhancement grayscale.This step aims to determine skin area in-ear image by space converting RGB applying certain rules area.the second applies segmentation split non-skin areas where detection.After stage, in edge is smoothing through a Gaussian filter then converted grayscale after an important ear.Finally Canny detection, general,...