Bloodless Technique to Detect Diabetes using Soft Computational Tool

Detection of diabetes using bloodless technique is an important research issue in the area of machine learning and artificial intelligence (AI). Here we present the working of a system designed to detect the abnormality of the eye with pain and blood free method. The typical features for diabetic retinopathy (DR) are used along with certain soft computing techniques to design such a system. The essential components of DR are blood vessels, red lesions visible as microaneurysms, hemorrhages and whitish lesions i.e., lipid exudates and cotton wool spots. The chapter reports the use of a unique feature set derived from the retinal image of the eye. The feature set is applied to a Support Vector Machine (SVM) which provides the decision regarding the state of infection of the eye. The classification ability of the proposed system for blood vessel and exudate is 91.67% and for optic disc and microaneurysm is 83.33%.

Exploiting Power Line for Communication Purpose

Power Line Communication (PLC) uses the available power line as a communication medium. The purpose of this chapter is to present the salient features, current trend and future scope of PLC with emphasis in the Indian context. Unlike other channels available, power lines are harsh media for data transfer and require efficient modeling and simulation techniques to propose and implement suitable mitigation schemes for achieving acceptable performance. Designed equipments have to adhere to strict mandates at the national and international levels to account for issues related to electromagnetic compatibility (EMC). In spite of this, PLC is expected to occupy an important place in the networking market in applications of smart grid and as a component of heterogeneous/hybrid communication system. The chapter is also backed by results from experiments carried on a typical power line in a test site with a presentation of noise, transfer characteristics, modeling and an estimate of the channel capacity.

A Review on Vision-Based Hand Gesture Recognition and Applications

In the present scenario, vision based hand gesture recognition has become a highly emerging research area for the purpose of human computer interaction. Such recognition systems are deployed to serve as a replacement for the commonly used human-machine interactive devices such as keyboard, mouse, joystick etc. in real world situations. The major challenges faced by a vision based hand gesture recognition system include recognition in complex background, in dynamic background, in presence of multiple gestures in the background, under variable lighting condition, under different viewpoints etc. In the context of sign language recognition, which is a highly demanding application of hand gesture recognition system, coarticulation detection is a challenging task. The main objective of this chapter is to provide a general overview of vision based hand gesture recognition system as well as to bring into light some of the research works that have been done in this field.

Design of a Power Aware Systolic Array based Support Vector Machine Classifier

This chapter presents a method for generating binary and multiclass Support Vector Machine (SVM) classifier with multiplierless kernel function. This design provides reduced power, area and reduced cost due to the use of multiplierless kernel operation. Binary SVM classifier classifies two groups of linearly or nonlinearly separable data while the multiclass classification provides classification of three nonlinearly separable data. Here, at first SVM classifier is trained for different classification problems and then the extracted training parameters are used in the testing phase of the same. The dataflow from all the processing elements (PEs) are parallely supported by systolic array. This systolic array architecture provides faster processing of the whole system design.

A New Coding Scheme for Data Security in RF based Wireless Communication

A radio-controlled (RC) aircraft is controlled remotely by a hand-held transmitter and a receiver within the craft. The working mechanism of such an arrangement designed using an AT89S51 microcontroller is reported in this chapter. The primary focus of the chapter is to describe the design of the interfacing of transceiver module with AT89S51 microcontroller and control the movement of the aircraft according to the instruction given remotely. The microcontroller reads the input given by the user and transmits the data to the receiver at the aircraft. The receiver module receives the transmitted signal and demodulates it and gives the data as serial sequence of bits at the output. The serial data are then given to the decoder which transforms the data from serial to parallel. This set of data is used to control motors and any related device. A special coding technique is used to secure the transmitted data.

Intelligent Fighter Pilot Support for Distributed Unmanned and Manned Decision Making

This chapter highlights important aspects of an intelligent fighter pilot support for distributed unmanned and manned decision making. First the background is described including current trends within the domain, and characteristics of a decision support system are discussed. After that a scenario and example situations are presented. The chapter also includes reflections of an intelligent fighter pilot support for distributed unmanned and manned decision making from the joint cognitive systems view, regarding human interoperability, and function allocation.

Acoustic Modeling of Speech Signal using Artificial Neural Network

Acoustic modeling of the sound unit is a crucial component of Automatic Speech Recognition (ASR) system. This is the process of establishing statistical representations for the feature vector sequences for a particular sound unit so that a classifier for the entire sound unit used in the ASR system can be designed. Current ASR systems use Hidden Markov Model (HMM) to deal with temporal variability and Gaussian Mixture Model (GMM) for acoustic modeling. Recently machine learning paradigms have been explored for application in speech recognition domain. In this regard, Multi Layer Perception (MLP), Recurrent Neural Network (RNN) etc. are extensively used. Artificial Neural Network (ANN)s are trained by back propagating the error derivatives and therefore have the potential to learn much better models of nonlinear data. Recently, Deep Neural Network (DNN)s with many hidden layer have been up voted by the researchers and have been accepted to be suitable for speech signal modeling. In this chapter various techniques and works on the ANN based acoustic modeling are described.

A Computer Based System for ECG Arrhythmia Classification

Biological signals can be classified according to its various characteristics like waveform shape, statistical structure and temporal properties. Among various bioelectric signals, one of the most familiar signal is the ECG. It is a signal derived from the electrical activity of the heart. The heart is an important organ which supplies body with oxygen. ECG is widely used in monitoring the health condition of the human. Cardiac arrhythmias can affect electrical system of the heart muscles and cause abnormal heart rhythms that can lead to insufficient pumping of blood and death risks. An important step towards identifying an arrhythmia is the classification of heartbeats. Modern analysis of electrical activity of the heart uses simple as well as sophisticated algorithms of digital signal processing. With the advent of technology, automatic classification of electrocardiogram signals through human-computer interactive systems has received great attention. This chapter discusses some computer assisted classification techniques based on statistical features extracted from ECG signal.

An EOG Signal based Framework to Control a Wheel Chair

Elecrooculogram (EOG) signal extraction is critical in the working of any electrooculography aided system based upon the tracking of the ocular movement of the eye dipole. In this chapter the signals captured using sensors (electrodes), are first amplified, then the noise is removed and then digitized, before being transferred to controller for movement of the wheelchair. Finally, from the muscle sensor, the output is directly being given to the controller to reach the target and complete the control of the movement of the wheelchair. Initially, a potentiometer is used instead of the Ag-Agcl electrodes to test the strength of signal obtained due to the movement of the eyes. Using this wheelchair is quite an advantage because this chair helps a physically handicapped person to move freely without being dependent on anyone else. The research provides a new method for human-machine interface system.

Utilizing UML, cTLA, and SRN

To meet the challenge of conducting quantitative analysis at the early stage of the system development process, this chapter introduces an extensive framework for performance modeling of a distributed system. The goal of the performance modeling framework is the assessment of the non-functional properties of the distributed system at an early stage based on the system's functional description and deployment mapping of service components over an execution environment. System's functional description with deployment mapping has been specified using UML. To analyze the correctness of the UML specification style, we have used temporal logic, specifically cTLA, to formalize the UML model. We have shown in detail how UML models are formalized by a set of cTLA processes and production rules. To conduct the performance evaluation of a distributed system, the UML model is transformed into analytic model SRN. We have specified an automated model transformation process to generate SRN model from UML, which is performed in an efficient and scalable way by the use of model transformation rules.