Abdolah Chalechale

Coin recognition using image abstraction and spiral decomposition

This paper presents a novel approach for coin image recognition. The approach enables measuring the similarity between full color multi-component coin images and needs no cost intensive image segmentation. A novel procedure, based on strong edges of the coin image, is exploited to derive an abstract image. Spiral decomposition of pixels in the abstract image is then used to extract a set of compact and effective features. The query set and the image database used in the tests are scanned, photographed, or collected from the web. The results are compared with three other well-known approaches within the literature. Experimental results show significant improvement in the Recall ratio using the proposed features.

Latency Optimization for Resource Allocation in Cloud Computing System

Recent studies in different fields of science caused emergence of needs for high performance computing systems like Cloud. A critical issue in design and implementation of such systems is resource allocation which is directly affected by internal and external factors like the number of nodes, geographical distance and communication latencies. Many optimizations took place in resource allocation methods in order to achieve better performance by concentrating on computing, network and energy resources. Communication latencies as a limitation of network resources have always been playing an important role in parallel processing (especially in fine-grained programs). In this paper, we are going to have a survey on the resource allocation issue in Cloud and then do an optimization on common resource allocation method based on the latencies of communications. Due to it, we added a table to Resource Agent (entity that allocates resources to the applicants) to hold the history of previous allocations. Then, a probability matrix was constructed for allocation of resources partially based on the history of latencies. Response time was considered as a metric for evaluation of proposed method. Results indicated the better response time, especially by increasing the number of tasks. Besides, the proposed method is inherently capable for detecting the unavailable resources through measuring the communication latencies. It assists other issues in cloud systems like migration, resource replication and fault tolerance.

Visual-based human-machine interface using hand gestures

This paper presents a new paradigm for visual-based interaction with computers using body gestures. The paradigm is based on statistical classification for gesture selection. It has applications in daily interaction with computers, computer games, telemedicine, virtual reality, and sign language studies. Specifically, hand gesture selection and recognition is considered as an example. The aims of this paper are: (a) how to select an appropriate set of gestures having a satisfactory level of discrimination power, and (b) comparison of invariant moments (conventional and Zernike) and geometric properties in recognizing hand gestures. Two-dimensional structures, namely cluster-property and cluster-features matrices, have been employed for gesture selection and to evaluate different gesture characteristics. Experimental results confirm better performance of the geometric features compared to moment invariants and Zernike moments.

A comprehensive mobile e-healthcare system

Bio-sensors and communication devices like smart phones witnessed significant progress, to take this advances, we propose a mobile heath-care system that reduce the distance between the patient and the health care center especially for the patient that need long term nursing care. This mobile system provide a good monitoring system for the patient in outdoor and indoor. The system is divided into two part, the first is a mobile phone application installed on a smart phone that connected with Bio-sensors, the second is a data base center that connected with different health-care center and then the ambulance center. The application record the Bio-signals from sensors and decide the patient status, send important information to the center, then add it to the patients history that can accessed by the physician from any where.

Coin image analysis, semantic evaluation and efficiency assessment

This paper presents a novel approach in coin image analysis based on middle-level features. The approach is used for image matching and it is developed utilizing edge pixel neighboring histograms. The overall structure of coin images are captured and used for similarity measurements. The closeness of the method to human perceptual judgment is evaluated by a semantic-based test along with the efficiency in comparison with three other methods known from the literature. The query set and the image database used in the tests are scanned, photographed, or collected from the Web. Experimental results show a significant supremacy of the new approach in semantic compatibility over the histogram of edge angles and the invariant moments. The proposed method is more efficient than the correlation method which is closer to human perception.

Parallelized computation for Edge Histogram Descriptor using CUDA on the Graphics Processing Units (GPU)

Most image processing algorithms are inherently parallel, so multithreading processors are suitable in such applications. In huge image databases, image processing takes very long time for run on a single core processor because of single thread execution of algorithms. GPU is more common in most image processing applications due to multithread execution of algorithms, programmability and low cost. In this paper we show how to implement the MPRG-7 Edge Histogram Descriptor in parallel using CUDA programming model on a GPU. The Edge Histogram Descriptor describes the distribution of various types of edges with a histogram that can be a tool for image matching. This feature is applied to search images from a database which are similar to a query image. We evaluated the retrieval of the proposed technique using recall, precision, and average precision measures. Experimental results showed that parallel implementation led to an average speed up of 14.74×over the serial implementation. The average precision and the average recall of presented method are 67.02% and 55.00% respectively.

Recognition of the triangular and rectangular road signs using passing lines intersection

A road sign recognition system automatically detects road signs as images captured by the camera installed in the car. It helps the driver to drive better. Most of the existing systems have two phases, including detection and classification. This paper presents Passing Lines Intersection as a innovative method for the classification phase. It is capable to classification the triangular and rectangular road signs swiftly and showing 89.26% accuracy rate. The main idea of the proposed method for road sign classification is to analyze special lines that are unique in all signs. A significant characteristic of the method is independence of the scale, direction and rotation of the road signs. Furthermore, the database used in the system can easily change for different countries having different signs.

Parallelization of a color DCT watermarking algorithm using a CUDA-based approach

Image watermarking in DCT domain has a high computational complexity especially for color and high resolution images, where usage of them has been significantly grown. To address this issue, in this article, a data-parallel color DCT watermarking approach is proposed and implemented on GPU using CUDA. Also, in this work, before embedding, the color watermark is compressed using a modified method to get less distortion. CUDA implementation of 8×8 DCT offers 12×-43× speedup with GT 540M and 94×-105× speedup with GTX 580, for different image sizes. In case of embedding procedure, the speedup obtained by GT 540M is between 7× and 26×, and the speedup obtained by GTX 580 is between 46× and 73×, for various case studies. Furthermore, in case of extracting procedure, GT 540M leads to a speedup between 10× and 29×, and GTX 580 leads to a speedup between 75× and 80×, for various case studies.

An Evolutionary Stochastic Approach for Efficient Image Retrieval using Modified Particle Swarm Optimization

Image retrieval system as a reliable tool can help people in reaching efficient use of digital image accumulation; also finding efficient methods for the retrieval of images is important. Color and texture descriptors are two basic features in image retrieval. In this paper, an approach is employed which represents a composition of color moments and texture features to extract low-level feature of an image. By assigning equal weights for different types of features, we can’t obtain good results, but by applying different weights to each feature, this problem is solved. In this work, the weights are improved using a modified Particle Swarm Optimization (PSO) method for increasing average Precision of system. In fact, a novel method based on an evolutionary approach is presented and the motivation of this work is to enhance Precision of the retrieval system with an improved PSO algorithm. The average Precision of presented method using equally weighted features and optimal weighted features is 49.85% and 54.16%, respectively. 4.31% increase in the average Precision achieved by proposed technique can achieve higher recognition accuracy, and the search result is better after using PSO.

A hybrid neural network using ICA and CGA for skin detection in RGB images

Skin color recognition is a useful and popular method in human-computer interaction and also in analyzing the content. In addition, the application programs for recognizing and detecting human body parts, faces, naked people, and retrieving individuals in multimedia databases all make use of skin recognition. Thus, finding a suitable method in order to segment the pixels of an image into different groups such as skin can be very important. Imperialist competitive algorithm (ICA) is a recently introduced evolutionary algorithm that showed a promising performance in some of the optimization problems. In this article, first the combined ICA-ANN, continuous genetic algorithm (CGA) and gradient descent algorithm were proposed and their performance was tested on images in RGB color spaces. In the proposed algorithms, a multilayer perceptron neural network manages the problems constraints, and ICA and genetic algorithms search to calculate the best response than the gradient descent algorithm. The proposed skin classification algorithms perform directly on the RGB color space. The results clearly indicate that the proposed algorithm significantly improves the performance of an MLP neural network.