Hasan Sarwar

FPGA realization of multipurpose FIR filter

A multipurpose FIR filter has been designed and realized by field programmable gate arrays (FPGA) for real-time filtering applications. The design can accomplish an arbitrary filter frequency response and variable filter order up to 127. The coefficients are computed through the Hamming windowing technique. The model is capable of performing filtering operations, like lowpass, highpass, bandpass and bandstop based on selection that is embedded into the design. The filter is set to 8-bit signed data processing. Linear constant coefficient difference equation (LCCDE) has been used to filter the input data in time domain. The design is coded with VHDL to cope with the parallelism of digital hardware. Simulation, compilation and synthesis have been done to verify the validity of the design outputs. To test the correctness of the design the observed output is compared with the calculated output results from MATLAB implementation that confirms the effectiveness of the design.

An Improved Image Denoising and Segmentation Approach for Detecting Tumor from 2-D MRI Brain Images

Image denoising and segmentation are the two most challenging fields in medical image processing particularly when it is application specific. The presence of noise not only degrades the visual quality but also immensely affects the accuracies of segmentation which is essential for medical diagnosis process. In this paper, we present an improved approach for denoising and segmentation of 2-D magnetic resonance brain images for detecting the tumor. We use fourth order partial differential equation based technique for removing MRI noises and thereby applied segmentation using automatic seeded region growing algorithm for detecting brain tumor automatically. The contribution of this work is the use of compass operator to preserve the anatomically significant information at the edges and a new morphological technique for skull removal from the brain MRI image which leads to the process of detecting tumor accurately. The method is tested on several real brain MRI images and it shows 100% success in detecting tumor automatically.

Prediction of State of Wireless Network Using Markov and Hidden Markov Model

Optimal resource allocation and higher quality of service is a much needed requirement in case of wireless networks. In order to improve the above factors, intelligent prediction of network behavior plays a very important role. Markov Model (MM) and Hidden Markov Model (HMM) are proven prediction techniques used in many fields. In this paper, we have used Markov and Hidden Markov prediction tools to predict the number of wireless devices that are connected to a specific Access Point (AP) at a specific instant of time. Prediction has been performed in two stages. In the first stage, we have found state sequence of wireless access points (AP) in a wireless network by observing the traffic load sequence in time. It is found that a particular choice of data may lead to 91% accuracy in predicting the real scenario. In the second stage, we have used Markov Model to find out the future state sequence of the previously found sequence from first stage. The prediction of next state of an AP performed by Markov Tool shows 88.71% accuracy. It is found that Markov Model can predict with an accuracy of 95.55% if initial transition matrix is calculated directly. We have also shown that O(1) Markov Model gives slightly better accuracy in prediction compared to O(2) MM for predicting far future.

Development of a Recognizer for Bangla Text: Present Status and Future Challenges

Optical Character Recognition (OCR) System, by virtue of its usefulness, has emerged as a major research area since 1950. Now it is becoming a more challenging issue all over the world to have efficient and more accurate recognizers. There are many widely spoken languages in the world like Chinese, Arabic, Hindi, English, Spanish, Bangla, Russian, Japanese etc. Bangla is one of the most widely spoken languages, ranking 5th in the world. 21st February is observed as the international mother language day to pay homage to the martyrs fought for the establishment of Bangla as the mother tongue of Bangladesh. With the automation everywhere, it is a burning issue to digitize huge, volume of Bangla documents by using an efficient OCR. However as of today there is no such good recognizer available for Bangla compared to other languages. From 80s, it took huge interest and now becomes as a major research area particularly in Bangladesh and India. Lots of works have been done in different sections of pattern recognition tasks (i.e, pre-processing, segmentation, feature extraction, classification) but there is a lack of synchronization between these works. That is why we put our effort into a comprehensive review of the current status of research to develop an all-inclusive Bangla OCR which will enable one to understand the difficulties and challenges involved, to know how much progress has been done and to estimate what more to be done to come out with a successful Bangla OCR.

An Algorithm for Segmenting Modifiers from Bangla Text

Script segmentation is an essential and preprocessing task for any OCR system. Bangla is one of the most popular scripts in the world. Since segmentation effects the recognition process, accurate and proper segmentation is necessary to implement Bangla OCR. Many works have been done for both handwritten and printed Bangla text. This paper presents the segmentation process of different Bangla modifiers from printed Bangla words.

Projection-based features: A superior domain for handwritten Bangla basic characters recognition

In spite of wide use of projection-based features in handwritten character recognition of several languages, its implementation was somewhat scanty in Bangla handwritten character recognition. This paper introduces the usage of projection profile features in recognizing handwritten Bangla basic characters. Alongside it also demonstrates a qualitative and quantitative analysis to visualize the effect of using projection based features on accuracy of recognition of Bangla handwritten characters through a number of approaches. In fact, this particular effort comprises of five different approaches where first one used longest-run, quad-tree and octant centroid features, second one adopted additional shadow features in association with the features of first approach, third one used longest run, quad-tree, shadow and chain code histogram features, next approach used longest-run, quadratic center of mass, shadow and left projection profile features and finally fifth approach with additional right projection profile features along with other features involved in the fourth approach. Throughout this analysis, neural network (trained via back-propagation algorithm) acted as classifier to observe the change in accuracy of recognition. It is seen that, with the increase in number of projection-based features, percentage of accuracy enhances at a greater rate than in case of inclusion of other features. This effective analysis can certainly assist a researcher to choose the optimal feature vector (consisting of several feature sets) for handwritten Bangla basic characters recognition.

An analysis of a videoconference-based teleconsultation system in Bangladesh.

In Bangladesh, telemedicine has been used at the Centre for the Rehabilitation of the Paralysed (CRP) since 1999. Elsewhere in Bangladesh it has been used by store and forward techniques, by SMS, by telephone and by videoconference. With the exception of the CRP telemedicine work, few of these attempts have continued beyond the pilot phase. We have studied the usage and acceptance of a videoconferencing system that is used by the Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders to provide telemedicine services in a distant district, Faridpur.

Analysis of real-time multimedia traffic in the context of self-similarity

Multimedia networks are now providing versatile services to meet different types of subscriber needs including voice, video, data, etc. Voice and video conferencing applications over these IP networks have already gained wide acceptance in todaypsilas end-user communities. The traffic characteristics and intensity as a function of time, geographic source can be realized by observing traffic distribution. Multimedia traffic has characteristics very different from Poisson characteristics, generating high rate data at one time and low rate data at another. Variable bit rate (VBR) real-time applications such as compressed video and audio and also the data sources tend to generate bursty traffic patterns and exhibit certain degree of correlation between arrivals and show long-range dependence in time i.e. self-similarity [1, 2, 3, 4, 5]. Therefore, it is important to identify and characterize network traffic flows to analyze network performance. In the present work the analysis of real time audio and video streaming traffic, colleted from the videoconferencing session of an enterprise high-speed hybrid multimedia network, has been done in the context of self-similarity.

A cost-effective web-based teleconsultation system

Common to other developing countries, telemedicine in Bangladesh has also not yet been widely used due to the expensive access to technology and lack of public awareness. In this paper, a simple and affordable web based teleconsultation system is introduced for remote areas where even basic healthcare is a critical issue. A web application has been developed using PHP and MYSQL for creating an asynchronous system. In this asynchronous case, the expert physicians have the flexibility to access medical data for examination at their convenience. The main goal of this paper is to build a cost effective system that can give easy health care access to rural peoples.

Prediction of the density of active wireless device using markov model

Location management is one of the key issues in wireless networks to provide an efficient and low-cost service. Realistic modeling of user mobility is a critical research area in wireless network. Mobility data based on real human behaviors may give us the opportunity to improve wireless and Mobile services for users in many ways. At present, several mobility models are proposed based on the analysis of real traces . In this paper, we investigate the feasibility of next state prediction using sequences of previously observed state and analyze the efficiency of MARKOV MODEL . The scenario concerns servicing wireless devices by wireless access point in the Dartmouth college campus over some period of time. The performance of the method has been verified for prediction accuracy. It is found that, on average, the choice of training data leads to prediction accuracy of 78.45%, in some cases the accuracy achieves about 95.55%.