Search

Cardiovascular disease (CVD) is one of the prime reason for death in India and across the globe. Rural areas of India suffer from shortage of cardiologist and medical facilities. Hence there is a need for the development of an efficient, automated heart disease detection system that can analyse the phonocardiogram to detect the disease. The paper proposes deep learning architectures for anomaly detection from heart sounds. The work classifies the unsegmented phonocardiograms into five classes, four cardiovascular diseases and normal(N). The detected pathological conditions are mitral valve prolapse (MVP), mitral stenosis (MS), mitral regurgitation (MR) and aortic stenosis (AS). Features are extracted using Mel Frequency Cepstral Coefficient (MFCCs) and learning and classification are performed using deep learning methods such as Convolutional Neural Network (CNN), Long Short Term Memory (LSTM) and a combination of 1DCNN and LSTM. A total of 1960 phonocardiogram (PCG) segments are used to develop the models with 392 segments in each class. We have achieved an accuracy of 99.1%, 98.2%, 99.4% for CNN, LSTM and 1DCNN-LSTM respectively.

Digital images have an inevitable role in almost all areas like clinical imaging, media broadcasting, crime analysis and scientific analysis etc. The phrase, “One image is equal to a thousand words”, is exactly fit as an image has more expressive capability than text messages. In all investigation, a photograph was always perceived as ‘proof of occurrence of an event’, due to a strong observation that, ‘seeing is believing ‘. Hence, images were considered as a ‘piece of truth’. Normally, an image is authentic if it is an originally recorded or captured from an actual scene or situation using any image capturing device. The captured image is expected to convey the original situation or scene at the source while capturing it in a real sense. In recent decades, due to the accessibility of modern photograph content altering software tools, it is very easy to transform the content of the image and therefore authenticity and integrity of the image is meager and tampering in digital images therefore does not require any specialist skills. This paper presents the modern methodological assessment and analysis of recent image forgery detection techniques. The various methods used in each stage of forgery detection techniques are also briefed. For an immediate reference, the tables of comparison are provided. The subject review paper is intended to help researchers to deliver useful understandings and modernized info about ongoing progress in forgery detection.

Air pollution has severe effects on human health and ecosystem. In India, air pollution levels are very high due to chemical gases and particulate matter emitting from automobiles, industry etc. Among the many pollutants PM 2.5 is considered as very crucial, as it can directly affect human health. Therefore, in recent years accurate prediction of PM 2.5 becomes a critical issue. In the present work, the air quality data from Thiruvananthapuram, Kerala has been used. Initially, data processing is done to fill missing values and for making the data stationary. Then data is normalized using Min-Max normalization. Finally, a LSTM model has been used to forecast daily PM 2.5 concentration. The results are evaluated using performance measures MAE and RMSE and is compared to a baseline persistence model.

Shock absorbers are used in vehicle to reduce the vibration and shocks while travelling. But during the transportation of sophisticated equipment, satellite launch vehicle parts and hazardous materials in goods carrying vehicles, more care is required. Special type of fixtures with vibration isolators is required for such vehicles in order to reduce the vibrations and thus ensuring smooth transportation. This work proposes a rugged sensor network-based vibration data acquisition system, which can be used during the transportation. Algorithms are developed for processing the data and to measure the vibration levels and the frequency. The developed system can also be used for the vibration analysis of passenger or goods vehicle.

This paper describes an automated medicine unit in which pills and medicines can be delivered at a lower cost than those provided at the medical stores. The implemented system also provides a centralized database which stores the medical history of each and every patient. It is possible to recover the data for further analysis and obtain the information of all the illness affected by the patient. In future, the system will be expanded to an ‘Anytime Medicine Kiosk’ where the common man can avail medicines at any time at a lower cost and the medical history of all the people in a given area can be obtained easily.

The Lensless Smart Sensor (LSS) developed by Rambus Inc. uses a revolutionary new approach to optical sensing by capturing information rich data in a tiny form factor. This paper experimentally verifies the superior quality of the new and improved version of the LSS as opposed to its earlier version. Also the paper examines the performance of the new sensor in 3D single point tracking. Sensor characteristics and accuracy are assessed which prove that the new sensor has considerable improvement and is able to track a point source down to millimeter level accuracy.

The Lensless Smart Sensor (LSS) developed by Rambus, Inc. is a low-power, low-cost visual sensing technology that captures information-rich optical data in a tiny form factor using a novel approach to optical sensing. The spiral gratings of LSS diffractive grating, coupled with sophisticated computational algorithms, allow point tracking down to millimeter-level accuracy. This work is focused on developing novel algorithms for the detection of multiple points and thereby enabling hand tracking and gesture recognition using the LSS. The algorithms are formulated based on geometrical and mathematical constraints around the placement of infrared light-emitting diodes (LEDs) on the hand. The developed techniques dynamically adapt the recognition and orientation of the hand and associated gestures. A detailed accuracy analysis for both hand tracking and gesture classification as a function of LED positions is conducted to validate the performance of the system. Our results indicate that the technology is a promising approach, as the current state-of-the-art focuses on human motion tracking that requires highly complex and expensive systems. A wearable, low-power, low-cost system could make a significant impact in this field, as it does not require complex hardware or additional sensors on the tracked segments.

Hand motion tracking traditionally requires highly complex and expensive systems in terms of energy and computational demands. A low-power, low-cost system could lead to a revolution in this field as it would not require complex hardware while representing an infrastructure-less ultra-miniature (~ 100μm - [1]) solution. The present paper exploits the Multiple Point Tracking algorithm developed at the Tyndall National Institute as the basic algorithm to perform a series of gesture recognition tasks. The hardware relies upon the combination of a stereoscopic vision of two novel Lensless Smart Sensors (LSS) combined with IR filters and five hand-held LEDs to track. Tracking common gestures generates a six-gestures dataset, which is then employed to train three Machine Learning models: k-Nearest Neighbors, Support Vector Machine and Random Forest. An offline analysis highlights how different LEDs' positions on the hand affect the classification accuracy. The comparison shows how the Random Forest outperforms the other two models with a classification accuracy of 90-91 %.

This paper presents the applicability of a novel Lensless Smart Sensor (LSS) developed by Rambus, Inc. in 3D positioning and tracking. The unique diffraction pattern attached to the sensor enables more precise position tracking than possible with lenses by capturing more information about the scene. In this work, the sensor characteristics is assessed and accuracy analysis is accomplished for the single point tracking scenario.

In the past decades satellite imagery has been used successfully for weather forecasting, geographical and geological applications. Low resolution satellite images are sufficient for these sorts of applications. But the technological developments in the field of satellite imaging provide high resolution sensors which expands its field of application. Thus the High Resolution Satellite Imagery (HRSI) proved to be a suitable alternative to aerial photogrammetric data to provide a new data source for object detection. Since the traffic rates in developing countries are enormously increasing, vehicle detection from satellite data will be a better choice for automating such systems. In this work, a novel technique for vehicle detection from the images obtained from high resolution sensors is proposed. Though we are using high resolution images, vehicles are seen only as tiny spots, difficult to distinguish from the background. But we are able to obtain a detection rate not less than 0.9. Thereafter we classify the detected vehicles into cars and trucks and find the count of them.

The development of high-resolution satellite imaging sensors has enhanced the collection of remotely sensed data and offers a cost-effective way to obtain information over large areas. The main structural features that are to be detected through satellite imaging are roads, buildings, and bridges. Detection of clouds and shadows supports the extraction of these features. However, the type of satellite images provided by different imaging sensors varies widely depending on spatial and spectral resolutions, which makes the process of feature extraction more complicated. This paper presents a detailed analysis of different algorithms developed to extract structural features using a variety of satellite images available for a single region. A comparative study of various techniques is done to justify the best method for the extraction of each feature.

Alzheimer's Disease (AD) is an irreversible neuro degenerative brain disorder. The progression of AD can be traced from Magnetic Resonance Images (MRI) and Positron Emission Tomography (PET) images. The cortical features can be well extracted from MR images while PET images clearly resolve the subcortical structures of brain. The registration approach used in this paper is a multistructure registration approach. The multistructure registration approach eliminates the problem of partial voluming and information loss. The segmentation of white matter and gray matter tissues traces the amount of tissue loss associated with AD. The multistructure registration of PET and MR images allows loading information in to a central space. By adopting multimodal neuroimaging the correlation between the modalities is exploited for the study. The registration procedure is based on the fluid kinematic model of the brain volume. This preserves the underlying topology and anatomy while constraining the transformation to be smooth. In this paper a novel framework for the early detection of AD using multimodal neuroimaging is proposed. The differentiation of AD from Mild Cognitive Impairment (MCI) was also studied.

Power Factor which is the ratio between the real or actual power and the apparent power is a very essential parameter in power system. It indicates how effectively the real power of the system has been utilized. In any electrical power system, a load with a low power factor draws more current than a high power factor load, for the same amount of useful power transferred. The most popular topology in Power Factor Correction (PFC) applications is certainly the boost topology. The boost topology is very simple and allows low-distorted input currents and almost unity power factor with different control techniques. A new fuzzy logic control strategy in a boost converter based PFC method for single phase rectifier is presented in this work. The proposed fuzzy logic control system has two inputs and one output. The proposed PFC control is based on boost converter operating at continuous conduction mode and provides a higher switching frequency.

Gyroscopes are used for maintaining orientation or measuring the angle of rotation or the rate of change of angular rotation. This paper compares angular rate measurement between MEMS digital (ADXRS 450) and analog (ADXRS 649) gyroscope. PIC 18F6520 is used to read the extracted angular rate from ADXRS 450. Digital output extracted from the sensor is error free even in noisy condition. Angular rate output from analog gyro can be read directly without the help of microcontroller but obtained output requires amplification. Difference amplifier AD8202 is used here for the amplification purpose. LabVIEW (Laboratory Virtual Instrumentation Engineering Workbench) software running on the computer will extract and display the data from the serial port at real time.

Vibrations produced by machinery are vital indicators of machine health. Vibration analysis is used as a tool to determine a machine's condition and the specific cause and location of problems, expediting repairs and minimizing costs. Machinery monitoring programs record a machine's vibration history. Monitoring vibration levels over time allows prediction of problems before serious damage can occur. Acoustic signal and vibration are correlated as pressure variations due to vibration create acoustic signals. The current trend of spacecraft design is to use complex and lightweight space structures to achieve increased functionality at a reduced launch cost. The goal of this research is to analyse output signals from suitable MEMS vibration as well as acoustic sensor ICs and to compare them with piezoelectric sensors. The study also aims in wireless monitoring of vibration as well as acoustic signals from rotating mechanical structures. MEMS sensors are low cost high resolution sensors with promising applications in low cost space applications such as Sounding rockets, Nano satellites and other terrestrial applications.

In this paper, a temperature measurement module using DS18B20 digital temperature sensor is developed to work as a standalone system. The module is compared with other conventional temperature sensors used in space applications. The total process consists of comparing the temperature obtained using DS18B20 module in real time with RTD, Thermocouple and LM35. PIC16F877A microcontroller is used to read the data from the sensor and send to PC. Lab VIEW (Laboratory Virtual Instrumentation Engineering Workbench) of National Instruments, running on the computer will extract and display the data from the serial port at real time. Finally, using BeagleBoard which is a low-power open-source single-board computer produced by Texas Instruments, the temperature is remotely monitored.

Satellite imaging sensors like SPOT, IKONOS and GEOEYE provide high resolution images, but the cost and complexity is high. If efficient algorithms exist for structural feature extraction like roads, buildings and bridges from satellite images, we can still use low resolution images to a certain extend. Many cases both low and high resolution images are affected by blurring and noisy effects, which seriously affects the feature extraction process. This paper provides an efficient algorithm for road network extraction from satellite images irrespective of their resolution, blurred and noisy effects.

Aerial and satellite images are projected images where clouds and cloud-shadows cause interferences in them. Detecting the presence of clouds over a region is important to isolate cloud-free pixels used to retrieve atmospheric thermodynamic information and surface geophysical parameters. This paper describes an adaptive algorithm to reduce both effects of clouds and their shadows from remote sensed images. The proposed method is implemented and tested with remote sensed RGB and monochrome images and also for visible (VIS) satellite imagery and infrared (IR) imagery. The results show that this approach is effective in extracting infected pixels and their compensation.

The problem of shadowing is particularly significant in high-resolution satellite imaging (HRSI) which causes either a reduction or total loss of information in the image. The research on segmenting shadow regions is of great significance for image interpretation and it is a supportive tool to detect manmade structures. This paper proposes a solution to the problem of automatic extraction of shadow features in color satellite images with high spatial resolution. The algorithm separates shadow from non-shadow by considering the spatial properties of the segmented regions. Experiments are executed on high resolution satellite images, and the results confirm the validity of the proposed method.

Automatic detection of artificial objects from satellite images are important source of information in many applications such as terrain mapping by remote sensing and GIS (Geographic Information System) applications. In this paper, a fuzzy based integrated algorithm for automatic detection of bridges over water is proposed. In the first step, the multispectral satellite image is given to a fuzzy based thresholding method to segment water regions from the background. Then candidate bridge pixels are extracted according to area analysis and bridge extraction algorithm developed. The algorithm is formulated in such a way that, the method can be applied to any complexity levels and any spatial resolutions. The approach in this paper has been implemented and tested with different types of satellite images to validate the superior performance of the algorithm.

Target tracking has a wide range of applications in Internet of Things (IoT), such as smart city sensors, indoor tracking, and gesture recognition. Several studies have been conducted in this area. Most of the published works either use vision sensors or inertial sensors for motion analysis and gesture recognition [1, 2]. Recent works use a combination of depth sensors and inertial sensors for 3D ranging and tracking [3, 4]. This often requires complex hardware and the use of complex embedded algorithms. Stereo cameras or Kinect depth sensors used for high precision ranging are instead expensive and not easy to use. The aim of this work is to track in 3D a hand fitted with a series of precisely positioned IR LEDs using a novel Lensless Smart Sensor (LSS) developed by Rambus, Inc. [5, 6]. In the adopted device, the lens used in conventional cameras is replaced by low-cost ultra-miniaturized diffraction optics attached directly to the image sensor array. The unique diffraction pattern enables more precise position tracking than possible with a lens by capturing more information about the scene.