Special Issue 2016, No. 6A
Print E-mail

De-Noise Steganography by Enhancing the Cover

Image:  Multi-Level Security Approach

 

 Jithesh Korothan1, S.B Kishor2 , and Pradeep Kumar3

1Department of Computer Science, Mahatma Gandhi College, India.

2Department of Computer Science, Sardar Patel Mahavidyalaya, India.

3Department of Computer Science, Kamala Nehru College, India.

Abstract: A technique that combines both logic and craft only can survive long. A multi-level security mechanism by blending steganogrpahy and visual cryptography is proposed here. Hiding already encrypted data inside an image is of immense value. Encryption is done by visual cryptography and hiding is done through the proposed de-noise Steganography With Selected Noise Bit Replacement Technique (SNBR). By and large, hiding data inside an image makes distortion to the image that leads to suspicions. Instead, this technique tries to remove disturbance or noise that is already present in the image and enhances the cover image through replacing noise bits present in the cover image with secret. In view of the fact that we replace the noise bits with the secret, it improves the quality of the cover medium and consequently the cover seems innocuous. As a result it reduces the chance of steg-analysis. The proposed method called SNBR, uses a location map to guarantee the correct extraction of the secret data. The goal of this study is to avoid degradation of the cover and improve the confidentiality of the information being communicated. Experimental results show that the new method achieves good security and a higher peak signal to noise ratio for the same number of bits per pixel of embedded image.

 

Keywords: Cover, de-noise, steganography, selected noise bit, stego-key, visual cryptography.

Received May 5, 2014; accepted December 23,2014

Full text  

 


 

 
Print E-mail

A Novel JFD Scheme for DRM Systems Based on DWT and Collusion Resistance Fingerprint Encoding

Reham Mostafa1, Hamdy El-Minir2 and Alaa El-Din Mohamed 1

1 Faculty of Computer and Information Science, Mansoura University, Egypt

2 Faculty of Engineering, Kafr El-Sheikh University, Egypt

Abstract: With the proliferation of the internet and rapid development of multimedia, media distribution and traitor tracing issues have become imperative and critical. In this paper, the Digital Rights Management (DRM) system based on novel Joint Fingerprinting and Decryption (JFD) scheme is proposed, which transmit the encrypted image to different customers and makes each customer decrypt the image into a different copy that contains the customer’s unique information. Till now, some JFD schemes have been reported, that solves encryption and fingerprinting simultaneously and has high efficiency, but several problems still remain to be tackled in JFD, including poor encryption security, severe fingerprinted image distortion, etc. An improved JFD scheme, which based on Discrete Wavelet Transform (DWT) of media and collusion resistant fingerprint encoding, is presented in the paper. The experimental results show that the proposed scheme is more secure compared with the existing scheme, it obtains good imperceptibility and robustness. These properties make it a suitable choice for secure image distribution in real time applications.

Keywords: DRM, JFD, traitor tracing, DWT, partial encryption.

Received April 30, 2014; accepted October 19, 2015

 

Full text 

 


 

 
Print E-mail

Classification of Carotid Artery Abnormalities in Ultrasound Images using an Artificial Neural Classifier

Dhanalakshmi Samiappan1 and  Venkatesh Chakrapani2

1Department of Electronics and Communication Engineering, SRM University, India.

2Department of Electronics and Communication Engineering, Sengunthar Engineering College, India.

Abstract: This work presents a computer-aided system for the identification of plaques and atherosclerosis of carotid abnormalities and the individuals at risk of stroke. Intima Media Thickness (IMT) of carotid artery is the standard biomarker of subclinical atherosclerosis and plaques. Conventional IMT measurement by expert sonologist is time consuming, associated with subjectivity and the process becomes difficult when the number of patients is very large. This paper proposes a standard protocol to diagnose patients efficiently and the process is made extremely fast. In this paper, the decision making ability of an artificial learning machine is investigated in carotid ultrasound artery image classification. Architecture with multilayer Back Propagation Network (BPN) using Levenberg-Marquardt training with good generalization capabilities and extremely fast learning capacity that overcomes the local minima problem of generalized BPN has been proposed. Carotid images are preprocessed, normalized and segmented to extract eighteen different feature sets and given as input to Artificial Neural Network (ANN). The selected features are found to be the good choice of feature vectors and have the ability to discriminate between normal and abnormal image. The proposed system is robust to any ultrasound image artifact. ANN classifier is evaluated using 361 ultrasound images. The efficiency is measured by validating the outputs of this decision support system with that of medical experts. This system improves the classification rate, reaching the diagnostic yield of 89.43%. The simulation results depicts that ANN achieves good classification accuracies with less implementation complexity when compared with manual operation.

Keywords: Artificial neural network, multilayer back propagation network, ultrasound carotid artery, carotid intima-media thickness, subclinical atherosclerosis, decision assist system.

Received November 8, 2013; accepted January 16, 2014

 

Full text  

 

 

 
Print E-mail

Unsupervised Feature Based Key-frame

Extraction Towards Face Recognition

Jana Selvaganesan1, Kannan Natarajan2

1Department of ECE, Mookambigai College of Engineering, India

2Department of CSE, Jayaram College of Engineering and Technology, India

Abstract: A convenient and most effective method of querying a video database for robust face recognition is by using key-frames extracted from the image sequence. In this paper we present a clustering based approach that bypasses the need for shot detection or segmentation, to extract the key-frames from the video using the local features, for the purpose of face recognition. Local features which are insensitive to noise, displacement, scale, rotation and illumination, are extracted from arbitrary points on the images based on Speeded Up Robust Feature (SURF) algorithm. The frames are then clustered using sequential K-means algorithm. A representative frame from each cluster called the key-frame is then determined for subsequent use in video based face recognition. The proposed method has been demonstrated with experimental results obtained using Honda/UCSD (name of a standard database available for face recognition research) dataset 1.

Keywords: Feature extraction, frame clustering, key-frame, SURF, unsupervised learning.

Received December 12, 2013; accepted January 27, 2015

 

Full text 

 

 
Print E-mail

Segmentation of Brain from MRI Head Images

Using  Modified Chan-Vese Active Contour

Model

Kalavathi Palanisamy and Somasundram Karuppanagounder

Department of Computer Science and Applications, Deemed University, India

Abstract: In this article, a new segmentation method to extract the brain from T1, T2 and PD-weighted Magnetic Resonance Image (MRI) of human head images based on Modified Chan-Vese (MCV) active contour model is proposed. This method first segment the brain in the middle slice of the brain volume. Then, the brain regions of the remaining slices are segmented using the extracted middle brain as a reference. The input brain image is pre-processed to find the rough brain. The initial contour for the MCV method is drawn at the center of the segmented rough brain image and is then propagated to reach the brain boundary. The result of this proposed method is compared with the hand stripped images and found to produce significant results. The proposed method was tested with 100 volumes of brain images and had accurately segmented the brain regions which are better than the existing methods such as Brain Extraction Tool (BET), Brain Surface Extraction (BSE), Watershed Algorithm (WAT), Hybrid Watershed Algorithm (HWA) and skull stripping using Graph Cuts (GCUT). 

 

Keywords:  Brain segmentation, skull stripping, brain extraction method, active contour, magnetic resonance image.

Received May 20, 2014; accepted September 9, 2014

 

 
Print E-mail

A DNA-Based Security solution Using Aggregated Chaos Cross and Cubic Map

Zeeshan Ahmad1, Hafiz Umar2, Chundong Li3 and Ling Chen3

1School of Electronic Engineering and Optoelectronic Technology Nanjing University of Science and Technology, China

2Department of Computer Science, Ghazi University, Pakistan

3College of Electronic and Information Engineering, Southwest University, 

Abstract: DNA, cryptography and chaos, can be combined as a whole aggregated DNA, chaos-based and cryptography Discrete Chaotic Cryptography (DCC) to encrypt and decrypt data simultaneously. A new image encryption scheme based on chaotic system and DNA encoding has been proposed in this paper. Firstly the image is permuted by mixing the horizontally and vertically adjoint pixels with the help of cross chaotic map. Afterwards, the permuted image is divided into least significant bits (LSB) and most significant bits (MSB).  Each LSB and MSB is further divided into two blocks and encoded by DNA sequence. The encoded blocks are combined and XOR operation is performed to get a diffused image. Finally the diffused image is permuted by a cubic chaotic map to accomplish the confusion phase and cypher image is obtained. Simulation results show that the proposed scheme can achieve good encryption and also provide resistance against different kind of attacks.

Keywords: Chaos Theory, DNA, Image Encryption, Cryptography.

  Received June 17, 2014; accepted August 16, 2015

 Full text 

 

 


 
Print E-mail

Multiple Class Image-based Vehicle Classification

using Soft Computing Algorithms

Sanjivani Shantaiya1, Kesari Verma2, Kamal Mehta3

1 Department of Computer Science and Engineering, CSV Techical University, India

2 Department of Computer Application, National Institute of Technology, India

3Department of Computer Science and Engineering, Nirma Univesity, India

Abstract: Automatic vehicle classification has expanded into a momentous topic of study due to its importance in autonomous navigation, traffic analysis, surveillance security systems, and transportation management. This paper presents multiclass image based classification of vehicles like bikes and cars using soft computing algorithms like artificial neural network decision tree and support vector machine as well. The objective of this paper is to automate the classification of vehicles from images. A good set of descriptor features that capture the most important properties of an object are used to identify the object uniquely. Different views of vehicle images are considered as important factor during classification process as six multiple class labels are created accordingly.

Keywords: Detection, classification, artificial neural network, decision tree, support vector machine.

Received April 25, 2014; accepted June 7, 2015

 


 
Print E-mail

Modified Image Segmentation Method based on Region Growing and Region Merging

Muthiah Mary1, Lekshmi Padma2 and Maria John3

1Faculty of Electronics and Communication Engineering, Noorul Islam University, India

2Department of Electrical and Electrnics Engineering, Noorul Islam University, India

3Electrical and Electronics Enhineering in St.Xavier’s Catholic College of Engineering, India

Abstract: Image segmentation is one of the basic concepts widely used in each and every fields of image processing. The entire process of the proposed work for image segmentation comprises of 3 phases: Threshold generation with dynamic Modified Region Growing phase (DMRG), texture feature generation phase and region merging phase. by dynamically changing two thresholds, the given input image can be performed as DMRG, in which the cuckoo search optimization algorithm helps to optimize the two thresholds in modified region growing. after obtaining the region growth segmented image, the edges are detected with edge detection algorithm. In the second phase, the texture feature is extracted using entropy based operation from the input image. In region merging phase, the results obtained from the texture feature generation phase is combined with the results of DMRG phase and similar regions are merged by using a distance comparison between regions. The proposed work is implemented using Mat lab platform with several medical images. the performance of the proposed work is evaluated using the metrics sensitivity, specificity and accuracy. the results show that this proposed work provides very good accuracy for the segmentation process in images.

 

Keywords: DMRG, edge detection algorithm, cuckoo search algorithm, texture generation, region merging.

Received September 10, 2014; accepted December 23, 2014

 
Print E-mail

Modified Texture, Intensity and Orientation

Constraint Based Region Growing Segmentation of

2D MR Brain Tumor Images

Angel Viji1 and J. Jayakumari2

1Department of Computer Science and Engineering, Noorul Islam University, India

2Department of Electronic and Communication, Noorul Islam University, India

Abstract: Image segmentation is a process of dividing an image into different regions such that each region is nearly homogeneous. Magnetic Resonance (MR) images always contain a significant amount of noise caused by operator performance, equipment and the environment which can lead to serious inaccuracies with segmentation. Radiologists perform diagnosis manually at early stage. It is a very challenging and difficult task for radiologists to correctly classify the abnormal regions in the brain tissue, because Magnetic Resonance Images (MRI) images are noisy images. Because the tumors are inhomogeneous, un-sharp and faint, but show an intensity pattern that is different from the adjacent healthy tissue, a segmentation based on intensity, orientation and texture properties is proposed here. With this approach the image segmentation problem can be formulated and solved in a principled way based on well-established mathematical theories. The image clustering using texture also reduces the sensitivity to noise and results in enhanced image segmentation performance. The ground truth of the tumor boundaries is manually extracted from publicly available sources. Experimental results show that our method is robust and more accurate than other well known models. The superiority of the proposed method is examined and demonstrated through a large number of experiments using MR images.

Keywords: Segmentation, MRI images, texture, region growing.

Received October 4, 2013; accepted April 28, 2014

 

 

 
Print E-mail

Comparing Performance Measures of Sparse Representation on Image Restoration Algorithms

Subramaniam Sakthivel1, Parameswari Marimuthu2 and Natarajan Vinothaa3

1Department of Computer Science and Engineering, Sona College of Technology, India

2 Department of Computer Science and Engineering, Dhirajlal Gandhi College of Technology, India

3 Department of Computer Science and Engineering, Sri Ramakrishna Engineering College, India

Abstract: Image restoration is a systematic process that regains the lost clarity of an image. In the past, image restoration based on sparse representation has resulted in better performance for natural images. Within each category of image restoration such as de-blurring, de-noising and super resolution, different algorithms are selected for evaluation and comparison. It is evident that both local and non-local methods within each algorithm can produce improved image restoration results based on the over complete representations using learned dictionary. The Gaussian noise is added with the original image and comparative study is made from the three different de-noising techniques such as mean filter, Least Mean Square (LMS) adaptive filters and median filters. The experimental results arrived from the filters are discussed for each model of the selected image restoration algorithms-locally adaptive sparsity and regularization, Centralized Sparse Representation (CSR), low-rank approximation structured sparse representation and non-locally CSR. A comprehensive study of this paper would serve as a good reference and stimulate new research ideas in Image Restoration (IR).

Keywords: IR, sparse representation, image de-blurring, locally adaptive sparsity, CSR.

Received February 22, 2014; accepted July 9, 2014

 

 
Print E-mail

Copy-Move Forgery Detection Using Zernike and Pseudo Zernike Moments

Khaled Mahmoud and Arwa Abu-AlRukab

Computer Science Department, Zarqa University, Jordan

Abstract: Despite the fact that images are a primary source of information, the rapid growing of tools that used to amendment images makes the reliability of the digital images in risk. Copy-Move forgery is one important method to forge an image; where part of the image is copied and pasted in another part of the same image. Regarding the related literature in this topic, many methods were developed to detect Copy-Move forgery; each method has its own strengths and weaknesses. In this paper, the capability and the efficiency of using Pseudo-Zernike Moment (PZM) and Zernike Moments (ZM) in detecting this type of forgery are tested. For evaluating the performance of these methods, comprehensive and authentic dataset is used for testing purposes. The results showed that both methods (PZM-based and ZM-based) are robust against blurring, noise adding, color reduction, brightness change, and contrast adjustments that may affect the image with an acceptable false match. However, rotated and scaled copied region still give weak results. Moreover, the PZM-based method is slightly faster and more accurate than ZM-based method.

Keywords: Digital forensics, copy-move forgery, moments, ZM, PZM

Received May 9, 2016; accepted June 29, 2016

 

Full text 

 

 
Print E-mail

Efficient Color and Texture Feature Extraction

Technique for Content Based Image Retrieval

System

Jayanthi Karuppusamy and Karthikeyan Marappan

Department of Electronics and Communication Engineering, Tamilnadu College of Engineering, India

Abstract: The future user needs in the field of multimedia retrieval is the focus of many research and development activists. It is empirically observed that no single algorithm is efficient in extracting all different types of images like building images, flower images, car images and so on. Hence, a thorough analysis of certain color, texture and shape extraction techniques are carried out to identify an efficient Content Based Image Retrieval (CBIR) technique which suits for a particular type of images. The extraction of an image includes feature description, index generation and feature detection. The low-level feature extraction techniques are proposed in this paper are tested on Corel database, which contains 1000 images. The feature vectors of the Query Image (QI) are compared with feature vectors of the database images to obtain Matching Images (MI). This paper proposes Fuzzy Color and Texture Histogram (FCTH), and Color and Edge Directivity Descriptor (CEDD) techniques which extract the matching image based on the similarity of color and edge of an image in the database. The Image Retrieval Precision value (IRP) of the proposed techniques are calculated and compared with that of the existing techniques. The algorithms used in this paper are Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and Fuzzy linking algorithm. The proposed technique results in the improvement of the average precision value. Also FCTH and CEDD are effective and efficient for image indexing and image retrieval.

Keywords: CBIR, IRP, FCTH, CEDD.            

Received January 11, 2014; accepted June 18, 2015

 

Full text 

 

 
Print E-mail

A Brain Abnormality Detection and Tissue

SegmentationTechnique by Using Dual Mode

Classifier

Angel Viji1 and Jayakumari Jayaraj2

1Department of Computer Science and Engineering, Noorul Islam University, India

2Department of Electronic and Communication, Noorul Islam University, India

Abstract: In the analysis of brain Magnetic Resonance Images (MRI), tissue classification is an important issue. Many works have been done to classify the brain tissues from the brain MRI. This paper presents a new technique to classify the brain MRI images and to perform tissue classification by using dual mode classifier. Initially, the brain MRI images are obtained from the brain databases and features such as covariance and correlation are calculated from the input brain MRI images. These calculated features are given to Feed Forward Back Propagation Neural Network (FFBNN) to detect whether the given MRI brain image is normal or abnormal. After detection, the resultant image is subjected to the segmentation process with the use of Optimized Region Growing (ORGW) technique to accomplish efficient segmentation. Following that, by utilizing Local Binary Pattern (LBP), texture feature is computed from the segmented brain MRI images. Then this texture feature is given as the input to the dual mode classifier which has two branches. One branch classifies the normal tissues such as Grey Matter (GM), White Matter (WM) and Cerebrospinal Fluid (CF) and the other branch classifies the abnormal tissues such as Tumor and Edema. The performance of our proposed technique is compared with other techniques such as Conventional Region Growing (RGW), and MRGW.

Keywords: ORGW, LBP, dual mode classifier, FFBNN, correlation, covariance.

Received December 24, 2013; accepted November 4, 2014

 

 
Print E-mail

 

Vehicle Classification System Using Viola

 

Jones and Multi-Layer Perceptron

 

Tarig Almehmadi1, Zaw Zaw Htike2

1 Department of Electrical Engineering, University of Malaya, Malaysia

2Department of Mechatronics, International Islamic University, Malaysia

Abstract: The automatic vehicle classification system has emerged as an important field of study in image processing and machine vision technologies’ implementation because of its variety of applications. Despite many alternative solutions for the classification issue, the vision-based approaches remain the dominant solutions due to their ability to provide a larger number of parameters than other approaches. To date, several approaches with various methods have been implemented to classify vehicles. The fully automatic classification systems constitute a huge barrier for unmanned applications and advanced technologies. This project presents software for a vision-based vehicle classifier using multiple Viola-Jones detectors, moment invariants features, and a multi-layer perceptron neural network to distinguish between different classes. The results obtained in this project show the software’s ability to detect and locate vehicles perfectly in real time via live camera input.

Keywords: Automatic vehicle classification, viola-jones detection, moment invariants, neural network.

 

Received October 4, 2014; accepted September 16, 2015

 

Full text 

 
Print E-mail

Image Segmentation for the Extraction of Face using Haar Like Feature

Hemlata Arunachalam and Mahesh Motwani

 Department of Computer Science and Engineering, Rajiv Gandhi Technological University, India

Abstract: The segmentation of an image for the extraction of face is complex task. This paper presents a method for segmenting image for the extraction of human faces. The method is based on Haar Like Features (HLF) and it starts with skin colour detection in an input image. Then skin region is further processed by finding connected components and holes. Each connected component is tested to extract eye like holes by finding circularity and area. Each eye like holes is tested by comparing correlation coefficient to confirm as eyes. If eye like features exist in the connected component then the rectangular box is drawn to enclose each eyes, nose and mouth like region based on the distance parameter between two eye like holes. Then HLF is detected by finding integral image. Based on the comparison of haar difference and test rules, the final verification of each connected component as face is done. The detected face is enclosed in rectangle box using distance parameter of the line between two eyes. The proposed method is tested on Bao face database and experimental results shows that the method is effective and achieves better accuracy of face detection and has low error rate as compared to Viola-Jones [13] and combining Haar feature and skin colour based classifiers [3].

Keywords: Skin color, connected component, holes, HLF, haar, integral image.

Received November18, 2013; accepted May 21, 2014

Full text 

 

 

 
Print E-mail

The Influence of Non Functional Characteristics in

Effort Estimation using Function Point Analysis

Senthilkumar Murugesan and Chidambararajan Balasubramanian

Department of Computer Science and Engeneering, Kamaraj University, India

Abstract: Software engineering is the discipline which paves the road map for the development of software within the given schedule and effort with the desired quality. The process begins with estimating the size, effort and time required for the development of the software and ends with the product. In most existing research on the effort estimation, the Function Point Analysis (FPA) method is used to estimate the effort, but not ensure the non- functional characteristics and quality factors of the project. In this paper, we study the uncertainty of effort estimation in the project and the impact of non functional characteristics in the effort estimation in detail. The refined model shows that the influence of non functional characteristics increases the accuracy of effort estimation in software project. By implications, the research suggests somewhat the impacts of non-functional characteristics in estimation is the most effective approach to improve the estimating accuracy may be to make estimators and developers more accountable in the software estimation.

Keywords: Software effort estimation, FPA, non functional characteristics, accuracy.

Received November 26, 2013; accepted December 23, 2014

 


 
Print E-mail

A Novel Technique to Improve Template Security

for Biometric Recognition

Bharathi Subramaniam and Sudhakar Radhakrishnan

Department of Electronics and Communication Engineering, College of Engineering and Technology, India

Abstract: This paper presents one of the new fusion methods for multimodal biometric recognition to improve biometric template security. In this methodology, the process involves embedding finger vein into the hand vein image. Here the necessary features are extracted from the preprocessed images and subsequently, we embed the binary finger vein into hand vein by the use of embedded fusion. Finally the matching score is found out with the help of neural network. The proposed technique is tested on the standard data bases of finger vein and hand vein. This method provides lower false acceptance rate and false rejection rate when compared with other techniques, indicating the effectiveness of the proposed system.

Keywords: Template security, finger vein, hand vein, embedded fusion, multimodal biometric recognition.

Received September 11, 2014; accepted December 16, 2014

 

Full text 

 
Print E-mail

 Semantic Middleware: Multi-Layer Abstract Semantics Inference for Object Categorization

Peng Liu, Zhipeng Ye, Wei Zhao and Xianglong Tang School of Computer Science and Technology, Harbin Institute of Technology, China

Abstract: In this paper, we present a hierarchical model, named as Multi-layer Abstract Semantics Inference (MASI), based on bag-of-visual-words (BoVW) to solve the problem of universal image categorization, including typical and zero-shot image categorization. An abstract hierarchical semantics learning method is proposed in the training step by extracting and selecting abstract visual words in a bottom-up way to train abstract semantic classifiers. For a testing image, its category is estimated layer-by-layer from top to bottom according to its corresponding hierarchical categories. Experimental results on popular image datasets have shown that the proposed method achieves better performance compared with traditional learning methods.

Keywords: Image categorization, zero-shot learning, semantic abstraction, BoVW.

Received November 11, 2014; accepted December 21, 2015

 
Print E-mail

 A Novel Biometric Based on ECG Signals and Images for Human Authentication

Mohamed Hammad, Mina Ibrahim and Mohiy Hadhoud

Faculty of computers and information, Menoufia University, Egypt

Abstract: This paper represents a complete system for using Electrocardiogram (ECG) images for human authentication. In this study, the proposed algorithm is divided into three main stages: Pre-processing stage, feature extraction stage and classification stage. A real database is used; it consists of 120 ECG images which are collected from 20 persons. The preprocessing stage is done on the ECG image. Preprocessing should remove all variations and details from an ECG image that are meaningless to the authentication method. In addition, this paper discusses briefly an extended version of work previously published on ECG feature extraction. In classification stage, Neural Network is used to make persons authentication. At the end, a system for real-time authentication is built. The proposed system achieves high sensitivity results for extracting ECG features and for human authentication.

Keywords: ECG image, human authentication and neural Network.

Received March 3, 2015; accepted April 26, 2015

 
Print E-mail

Image Zooming Technique Based on the Split Bregman Iteration with Fractional Order Variation Regularization

Liping Wang, Shangbo Zhou and Awudu Karim College of Computer Science, Chongqing University, China

Abstract: It is always a challenging work to develop an accurate and effective method to reconstruct a degraded image. In this paper, the nonlocal variation Fractional Total Variation (FTV) regularization technique for image zooming is investigated. To enhance edges, yet preserve textures, fractional order calculus based image zooming method is proposed, which can deal well with fine structures like textures. To solve the nonlinear Euler-Lagrange equation associated with the nonlocal variation FTV regularization model, we propose a nonlocal total variation method for image zooming based on the split Bregman iteration. Enlarging and de-noising experimental results show that the proposed method has effectiveness and reliability by comparing to some methods mentioned in the paper.

Keywords: Image zooming, total variation, split bregman iteration, fractional order.

Received December 26, 2014; accepted June 1, 2015

Full text


 
Print E-mail

A Robust and Efficient Anti Spoofing Method for

Facial Recognition Systems using the Fusion of

Fresnel Transform and Micro-Texture Analysis

Farhood Mousavizadeh1, Keivan Maghooli1, Emad Fatemizadeh2 and Mohammad Moin3 1Department of Biomedical Engineering, Islamic Azad University, Iran 2School of Electrical Engineering, Sharif University of Technology, Iran 3Faculty of IT, ICT Research Institute (Iran Telecom Research Center), Iran

Abstract: Face biometric systems provide automatic verification or identification of a person. But nowadays using hacked or stolen photographs or videos is one of the most common manners for spoofing such systems. This problem can be solved by using some specific hardware’s like IR or stereoscopic cameras. However, the additional hardware should be low cost and applicable for the facial recognition purposes. To deal with the spoofing problem, we present single image and real-time method that can work with conventional cameras. Facial images commonly contain surface textures and the dept characteristics that cannot be found in a photograph and also there are some differences in the frequency distribution of a real face and a fake one. These two properties are the basic features of the most liveness detection systems. In this paper, we aim to utilize an automatically facial liveness detection method that combines these two features to have a robust and reliable method for single image liveness detection. We use the fusion of the Zernike moments of Fresnel transformed images and multi-scale Local Binary Patterns (LBP) histogram and fed them to Principal Components Analysis (PCA) and Fisher’s Discriminant Ratio (FDR) analyzers to obtain efficient and rich sets of features. The results show that we can achieve to the features that are half/quarter the size of original feature sets using FDR /PCA respectively. The results show that we could have liveness detection features stronger in performance and smaller in dimension in comparison with the common and state-of-the-art methods like LBP.

Keywords: Liveness detection, fresnel transform, local binary patterns, zernike moments analysis, FDR, PCA.

Received July 28, 2014; accepted May 11, 2015

 
Print E-mail

 Automated Nuclei Segmentation Approach based on Mathematical Morphology for Cancer Scoring in Breast Tissue Images

Aymen Mouelhi1, Mounir Sayadi1, Farhat Fnaiech1 and Karima Mrad2

1Laboratory of Signal Image and Energy Mastery, ENSIT-University of Tunis, Tunisia.

2Morbid Anatomy Service, Salah Azaiez Institute of Oncology, Tunisia

Abstract: In this work, we propose an automated approach able to perform accurate nuclear segmentation in immunohistochemical breast tissue images in order to provide quantitative evaluation of estrogen or progesterone receptor status that will help pathologists in their diagnosis. The presented method is based on color deconvolution and an enhanced morphological processing, which is used to identify positive stained nuclei and to separate all clustered nuclei in the microscopic image for a subsequent cancer scoring. Experiments on several breast cancer images of different patients admitted into the Tunisian Salah Azaiez Cancer Center, show the efficiency of the proposed method when compared to the manual evaluation of experts. On the whole image database, we recorded more than 97% for both accuracy of detected nuclei and cancer scoring over the truths provided by experienced pathologists.

Keywords: Breast cancer, immunohistochemical image analysis, color deconvolution, morphological operators.

Received September 11, 2014; accepted March 23, 2015

 
Print E-mail

 BGLBP-based Image Background Extraction

Method

Seyed Davarpanah1, Fatimah Khalid2 and Lili Abdullah2

1Department of Computer Engineering, Faculty of Engineering, University of Science and Culture, Iran

2Faculty of Computer Science and IT, Universiti Putra Malaysia, Malaysia

 

Abstract: Local Binary Pattern (LBP) is invariant to the monotonic changes in the grey scale domain. This property enables LBP to present a texture descriptor that can be useful in applications dealing with the local illumination changes. However, the existing versions of LBP are not able to handle image illumination changes, especially in outdoor environments. These non-patterned illumination changes disturb performance of the background extraction methods. In this paper, an extended version of LBP which is called BackGround LBP (BGLBP) is presented. BGLBP is designed for the background extraction application but it is extendable to the other areas as a texture descriptor. BGLBP is an extension of Direction (D-LBP), Centre-Symmetric LBP (CS-LBP), Uniform LBP (ULBP), and R-LBP and it has been designed to inherit the positive properties of previous versions. The performance of BGLBP as a part of background extraction method is investigated. 

 

Keywords: Texture descriptor, LBP, background extraction.

 

Received  September 10, 2014; accepted September 20, 2015

 Full test

 

 

 
Print E-mail

A Robust DHWT Based AES Encrypted Image Watermarking Scheme

Maheswari Sureshbabu1, Rameshwaran Kalimuthu2 and Chandra Vadivel3

1Department of Electrical and Electronics Engineering, Kongu Engineering College, India

2Department of Electronics and Communication Engineering, Shanmuganathan Engineering College, India

3Department of Electronics and Communication Engineering, Institute of Engineering and Technology, India

Abstract: In this paper, we propose the blind watermarking algorithm based on Double Haar Wavelet Transform (DHWT) for copyright protection of encrypted images. Watermark embedding is performed in wavelet transform domain. DHWT is applied to the original cover image and binary watermark. Selected subband is encrypted by Advanced Encryption Standard (AES) encryption scheme. Singular Value Decomposition (SVD) is applied on selected subband of both the cover image and binary watermark. Eigen values of selected subband of cover image are modified by the Eigen values of the selected subband of binary watermark. Experimental results show that the proposed algorithm achieves very high imperceptibility which is evidenced by high Peak Signal to Noise Ratio (PSNR) value for various gray scale encrypted images. Also, it produces very high robustness against various types of image processing attacks.

Keywords: DHWT, AES encryption, SVD, digital watermarking.

Received October 28, 2013; accepted May 15, 2014

 Full text


 
Print E-mail

Motion Estimation in Video Coding using Simplified Optical Flow Technique

Eskinder Ayele and Sanjay Dhok Center for VLSI and  Nanotechnology, Visvesvaraya National Institute of Technology, India

Abstract: We propose an interpolation­free sub­pixel motion estimation technique that particularly aims at providing accurate motion vectors, where conventional sub­pixel interpolation motion estimation algorithms used in video coding are too complex and time consuming. The proposed algorithm is a combination of block matching algorithms and simplified optical flow, which is Taylor approximation. The technique does not require any pixel interpolation and it is much faster than conventional motion estimation methods. Statistical results illustrate that the new technique performs quickly and accurately with a compatible performance with respect to the benchmarking full search algorithm.

Keywords: Sub­pixel motion estimation, block matching, interpolation, interpolation­free, optical flow.

Received November 11, 2013; accepted November 25, 2014

 
Print E-mail

Efficient Block-based Motion Estimation

Architecture using Particle Swarm Optimization

 

Vani Rajamanickam1 and Sangeetha Marikkannan2

1Meenakshi College of Engineering, Anna University, India

2Karpaga Vinayaga College of Engineering and Technology, Anna University, India

 

Abstract: High speed video transmission is the key to achieve high quality live or through offline streaming. Block matching motion estimation is adopted in video coding standards to improve the performance in terms of speed and at the same time, the power consumption should be minimal. The paper proposes an efficient Block-based Motion Estimation architecture, in which the Motion Vectors (MV) are obtained by searching for the best match in the previous frame. A resizable smart snake order is utilized for scanning the frames of different block sizes which improves the data reuse efficiency. The architecture is based on applying the global search ability of Particle Swarm Optimization (PSO) that reduces the number of logic elements. The parallel execution involved in the processing of sub-regions in the search window enables the architecture to achieve high speed. The proposed work coded in Verilog Hardware Description Language, and implemented with Altera Cyclone II FPGA, operates at a maximum frequency of 265.01MHz. It is observed that the total thermal power dissipation is 74.27 mw, making it suitably efficient for low power implementation of Motion Estimation.

 

Keywords: Resizable smart snake scan, swarm optimization, motion estimation, block matching.

 

Received October 18, 2013; accepted June 9, 2014

 Full text

 
Print E-mail

Extreme Curvature Scale Space for Efficient

Shape Similarity Retrieval

Hassan Silkan1, Said Ouatik2, and Abdelmounaime Lachkar2

1Department of Computer Science, Chouaib Doukkali University, Morocco

2Department of Electrical and Computer Engineering, Sidi Mohmmed Ben Abdellah University, Morocco

Abstract: The description of the object shape is an important characteristic of the image; several different shape descriptors are used. This paper presents a novel shape descriptor which is robust with respect to noise, scale and orientation changes of the objects. It is based on the multi scale space approach to identify shapes. The descriptor of a shape is created by tracking the position of extreme curvature points in a shape boundary filtered by low-pass Gaussian filters of variable widths. The result of this process is a several contours map representing the extreme curvature points of the shape as it is smoothed. The maxima of these contours are used to represent a shape. We demonstrate object recognition for three data sets, a classified subset of database SQUID, the set of silhouettes from the MPEG-7 database and the set of 2D views of 3D objects from the Columbia Object Image Library (COIL-100) database. The results prove the performance and robustness of the developed method and its superiority over Curvature Scale Space (CSS) in shape with shallow concavities.

Keywords: Multi scale analysis, CSS, shape similarity, image database retrieval.

Received January 18, 2014; accepted October 26, 2014

 

Full text 

 

 
Print E-mail

Cascading Deformable Parts Model in the Facial

Feature Detection of Frontal and Side Profile Images

Pisal Setthawong and Vajirasak Vanijja

School of Information Technology, King Mongkut's University of Technology, Thailand

Abstract: Facial feature detection is considered an important computer vision task that is used for many real world applications. Current advancements in computer vision have come up with many proposed facial feature detection approaches, such as Deformable Parts Model (DPM), that provide good accuracy in the detection of key facial features, but mainly in frontal poses. When presented with side profile poses many approaches do not perform well as certain facial features can be obscured and the approach may attempt to over-fit the trained model which leads to inadequate results. The proposal of a cascading pipeline extension to the DPM approach and a modified DPM approach for side profile specific facial feature detection is presented to deal with a wider range of facial profiles. The proposed approach would be evaluated empirically showing the improvement of the proposed method over a wide range facial feature configurations including side and frontal profiles.

Keywords: Facial feature detection, DPM, geometric model, image processing.

Received July 15, 2014; accepted March 2, 2015

Full Text

 

 

 

 
Print E-mail

Video Completion Using Gray Level Co-occurrence Matrix and Object Update

Sameh Zarif 1, 2, Ibrahima Faye 1, 3, and Dayang Rohaya 1, 2

1Centre of Intelligent Signal and Imaging Research, Universiti Teknologi Petronas, Malaysia

2Department of Computer and Information Sciences, Universiti Teknologi Petronas, Malaysia

3Department of Fundamental and Applied Sciences, Universiti Teknologi Petronas, Malaysia

Abstract: Reconstructing and repairing damaged parts after object removal of digital video is an important trend in artwork restoration. Video completion is an active subject in video processing, which deals with the recovery of the original data. Most previous video completion approaches consume more time in extensive search to find the best patch to restore the damaged frames. In addition to that, visual artifacts appear when the damaged area is large. This paper presents a video completion method without the extensive search process. The proposed framework consists of a segmentation stage based on low resolution version and background subtraction, a tracking stage based on Gray Level Co-occurrence Matrix (GLCM), and a completion stage based on object prior position and object update. The proposed method reduces the completion time to a few seconds and maintains the spatial and temporal consistency. It works well when the background has clutter or fake motion, and it can handle changes in object size and in posture.

Keywords: Video completion, video inpainting, object removal, GLCM, background subtraction.

Received March 21, 2014; accepted February 4, 2015

 
Print E-mail

Navigational Pattern Based Relevance Feedback Using User Profile in

CBIR

Syed Karim, Muhammad Harris, and Muhammad Arif

Department of Computer Science, City University of Science

and Information Technology Peshawar, Pakistan

Abstract: Content Based Image Retrieval (CBIR) is an application of computer vision and addresses the problem related to retrieval of digital images in large image databases. CBIR uses low level image features for retrieval task and tries to portray users intended results. Relevance Feedback (RF) is a technique for marking retrieved results as relevant or irrelevant by the user. People in the society have mutual interests and needs while searching for required data. Interesting and similar patterns can easily be found in the browsing behaviour of users pursuing required images from CBIR system. Recording users browsing behaviour and applying mining techniques to find frequent itemsets helps boost the retrieval performance of the CBIR system in terms of quality and processing time. User categorized into different groups on the basis of users age and gender specification helps fasten the mining process because of the similarity of thoughts in these users groups. This paper focuses on mining user browsing behaviour belonging to different user categories (user profiling) with FP-growth mining algorithm for revealing similar search patterns. The results show efficiency against the existing approach.

Keywords: CBIR, profile, apriori, FP-growth, support vector machine, navigational pattern relevance feedback.

Received June 15, 2014; accepted October 26, 2014

 
Print E-mail
 

A Secured Representation of Data Hiding in Wavelet Transform Domain

using Video Steganography

Mritha Ramalingam and Nor Isa

School of Electrical and Electronic Engineering, Universiti Sains Malaysia, Malaysia

Abstract: Hiding data in digital multimedia pertaining to text, image, audio and video plays a significant role in current trend for providing secured communication. Video Steganography is one of the emerging methods of sharing secret data by concealing the existence of data inside video files. The greatest advantage of using video for Steganography is that the large amount of data can be hidden into videos. Securing hidden data and minimizing the distortions to improve video quality after hiding data are the most important criteria which need to be considered as competing goals while designing a video Steganography system. A Secured Representation of Data Hiding (SRDH) is presented to enhance the security of hidden data using discrete cosine transform (DCT) coefficients of compressed video sequences as well as preserving the quality of video after hiding data using secured discrete wavelet transform (DWT). Video steganalysis is also examined to identify the videos containing hidden data. The performance of the proposed method is evaluated in terms of data embedding capacity, quality of service in terms of attack rate and noise level. Experimental results reveal the proposed method is more secure and produced better video quality as compared to existing algorithms.

Keywords: Video steganography, data hiding, discrete cosine transform, security, secured discrete wavelet transform.

Received April 16, 2014; accepted December 2, 2014

Full Text

 


 
Print E-mail

Improved Gaussian Mixture Model with Background Spotter for the

Extraction of Moving Objects

Brahim Farou1,2, Hamid Seridi2, and Herman Akdag3

1Computer Science Department, Badji Mokhtar-Annaba University, Algeria

2LabSTIC, Guelma University, Algeria

3LIASD, Paris 8 University, France

 Abstract: Extraction of moving objects is a key step in a visual surveillance area. Many background models have been proposed to resolve this problem, but Gaussian Mixture Model (GMM) remains the most successful approach for background subtraction. However, the method suffers from sensitivity (SE) to local variations; variations in the brightness and background complexity mislead the process to a false detection. In this paper, an efficient method is presented to deal with GMM problems through improvement on updating selected pixels by introducing a background spotter. First, the extracted frame is divided into several equal size regions. Each region is assigned to a spotter who will report significant environment changes based on histogram analysis. Only parts reported by spotters are considered and updated in the background model. Tests carried out on four video databases that take into account various factors, demonstrate the effectiveness of our system in real-world situations.

 Keywords: Video surveillance, GMM, modeling the background, image processing.

 

Received March 10, 2014; accepted December 23, 2014

Full Text

 
<
 
Copyright 2006-2009 Zarqa Private University. All rights reserved.
Print ISSN: 1683-3198.
 
 
Warning: fsockopen(): php_network_getaddresses: getaddrinfo failed: Name or service not known in /hsphere/local/home/ccis2k/ccis2k.org/iajit/templates/rt_chromatophore/index.php on line 251 Warning: fsockopen(): unable to connect to oucha.net:80 (php_network_getaddresses: getaddrinfo failed: Name or service not known) in /hsphere/local/home/ccis2k/ccis2k.org/iajit/templates/rt_chromatophore/index.php on line 251 skterr