Andrews Samraj

Emergency communication interface design using wearable data gloves for weary patients

The communication between care takers and patients are limited and complex when the patients are too weary and weak. A clinical base patient monitoring involves reading implicit communications of such patients carefully from their gestures to identify their need. The sensitivity of such system should be of high precision to avoid any misinterpretation that may lead to adverse effects. The wearable system can accurately interpret the implied communication to the care takers or to an automated support device. The simplest form of palpable hand movements are used for the above purpose. The proposed system suggests not only a novel methodology simpler than the existing sign language interpretations for such implicit communication, but also suggests the gestures appropriate to be used in such systems. The experimental results by two different modeling methods show a well-distinguished realization of diverse hand movement activities using a wearable sensor medium and the interpretation results always show significant and acute thresholds.

A segmented-mean feature extraction method for glove-based system to enhance physiotherapy for accurate and speedy recuperation of limbs

It is always desirable to have an accurate system that allows fast recovery of patients undergoing physiotherapy in terms of integrated health and cost benefits. The caregivers and medical personnel too gain a lot of expertise through the innovations involved in treatment methodology. This system proposed here was developed successfully with a straightforward Segmented-Mean feature construction method that enables its portability to suit smart biomedical devices. In this work, four different exercises were completed by four different subjects in two sessions and the feedback system was generated from every single trial performance via a visual display in a smart phone. The accuracy of the systems output depends on the precise representation of two important things namely, correct gesture and timings. These two parameters have to be captured from the signals that are generated by the hand glove during the manual physiotherapy as guided by the experts during the teaching (i.e. training) phase. Any deviation from the model should also be captured and reflected in the feedback to align the physio-movements towards perfection to minimise adverse effects. So the feature has to be constructed with complete representation and obviously, as fast as possible. The proposed Segmented-Mean method calculates the mean of data that arrives from the significant electrodes periodically, thus preserving the performance of the subject and is found suitable in estimating the enactment of exercises and required deviations (if any), accurately and as appropriate. The proposed Segmented-Mean method helps the construction of features easily than other conventional methods by reducing the computational complexity and therefore, the response time. Hence, shifting the importance to actual physiotherapy monitoring with an accurate system that works on simple feature construction made feasible.

Random Average Distribution aided SVD feature selection and interpretation for touch less robotic control by furtive gestures

The implicit robotic communication is highly appreciated in the fields of security and defense. The reason for such communication is to conceal the commands or its source to others to gain the whole benefits of the hidden robotic assistance to make it available in the case of emergency and critical situations. To interpret and communicate most distinctive orders to the robotic system, a well defined and distinguishable gesture paradigm is necessary. Moreover the conversion of such paradigms should also be precise and augmentative. The Random Average Distribution helps to reduce the complexity of the SVD features and augment the clear classification. The signals acquired from a pair of different gestures from five subjects with ten trials, were subjected to the proposed feature extraction method and were classified. The results obtained were found better than the results obtained by the direct SVD and reduced in complexity.

Communication by gestures in personal emergency response system

The Present systems on emergency response are primitive and not accurate in expressing the actual need of the person in risk. In a proper healthcare monitoring every communication between the patient and the care takers are critical and such response involves many implicit meanings. Even a slight misapprehension leads to adverse effects. A simple wearable system can precisely interpret the implicit communication to the care takers or to an automated support device. Simple and obvious hand movements can be used for the above purpose. The proposed system suggests a novel and swifter methodology simpler than the conventional sign language interpretations for such implicit communication. The exploratory experimental results show a well-distinguished realization of different hand movement activities using a wearable sensor medium and the interpretation results always show significant thresholds as well as faster recognition.

Identification of weed growth and intrusion in plant beds by modified singular value decomposition of areal sensory images

Plant monitoring is being done under precession agriculture now a days using areal sensors the segmentation of plant and decease symptoms in leaf image can be a valuable and timely aid for controlling the disasters early. In this research digital image of different plants along with leaf and other components have been segmented and quantified using simple techniques like SVD, modified methods on SVD feature construction. Our proposed approach helps to identify differences in plants much wider than conventional applications of SVD features. This threshold augmentation technique helps to widen the inter average differences from 7.20 to 8.08 where as the conventional method produce only a differences of 1.47 to 3.79, which is 3.44 times lower on an average than the proposed method.

Sign language communication and authentication using sensor fusion of hand glove and photometric signals

This paper proposes a sign language interpretation system by a combination of data glove and photoplethysmograph sensor measurements collected concurrently from data glove and PPG devices. This system is tested on many numbers of trials by different subjects performing a set of particular shapes using the American Sign Language (ASL). The sign language being the natural way of communication for the hearing and speech impaired people, there are systems being developed to understand the sign language and to produce communication. As an enhancement to the communication system for the targeted audience, authentication systems that can understand sign language would be of a great help. The focus of this paper is on the avoidance of skilled forgers try to seek the genuine sign language symbols in many numbers of trials. The reinforcement of sign language signals by the Photoplethesomograph signals were found to be increasing the gap in the Euclidian distances between forgers and the genuine template feature, thus prohibiting them from thrive forging. It has been clearly proved by our repetitive experiments here, on various subjects using the combinational features. The cause of benefits in the proposed system in eliminating forge attempts is due to the biometric differences found in data glove signal characteristics and level of bilateral hand dimensional dissimilarities plus arterial blood pulsation of individuals. Furthermore, we found some predominant sensors essential in this authentication process and trying to eliminate un useful sensors which are of less importance in hand glove thus making the optimization of hardware.

Weed growth and intrusion detection by multi angle images of vegetable plants using modified SVD

The requirement of earlier identification of weed growth and plant disease is one of the most important procedures in precision agriculture. In this research work we identified the best method of positioning the optical sensors to get the best possible images. To address this purpose the variant of SVD is used for feature construction. Our proposed approach optimize and reiterate our similar earlier findings in identification of plants. The threshold augmentation technique is used to find the inter and intra average differences for both aerial and portrait angle images for comparison. The intrusion detection software Matlab is created and tested for the Portrait images and compared with aerial images. The Portrait images to find the intrusion are found better than aerial images in number of trails 15% more than the aerial images.

Tool flank wears estimation by simplified SVD on emitted sound signals

A non disruptive Monitoring of tool flank wear is a prudent way in the Production process without affecting the throughput. The machining process which involves the tool wear condition inspection without stopping the process has to be done for this. A simple method is proposed through this paper with a novelty of reducing the processing time by sampling method to find the degree of wear of tool flank by closely analyzing a small portion of the sound signal emitted while the turning process captured and in progress. In this work an aluminium portion was used as a primary material on which the carbide material is used as the cutting edge. A uniform Cutting speed is maintained throughout the experiments. The constant feed rate and fixed cutting depth were maintained were as the flank wear assigned as floating. Recordings were done for the emitted sound signals when the turning process was in progress using a fresh tool (0 flank wear) then with another carbide tool with slight wear (0.2–0.25 mm) and finally with the same process but with a severely warned tool bit (0.4mm and above). These recordings were done using a high position microphone in an interleaved way. Individual analysis of the sound signals using Singular Value Decomposition (SVD) is done on pieces of recorded sound waves. A good correlation was formed with the increase in flank wear of the tools and the SVD values found for the pieces of sound waves irrespective of temporal intervals. Hence it can be decided that tool wear monitoring of flank while turning was in progress in small intervals by utilizing SVD features on such emitted sound signal is a prudent and significantly simple method and proved using iterative experiments.

Fuzzy ARTMAP based feature classification for danger and safety zone prediction for toddlers using wearable electrodes

The desired performance of every childcare and monitoring system is to clearly read the user activity into a relevant category of the solution domain. This categorization highly depends on error free processing methods and systematic regression or classification. The wearable interface acquires multiple signals of the user activity that serves as the input to the monitoring system. The pattern of the signal array after necessary consolidation and feature processing, determines its candidature into defined classes. Hence it is crucial to deploy a strong classifier which can characterize the activity of the user into normal zone activities or dangerous. In this paper, we used the robust and adroitness classification model Fuzzy ARTMAP to classify signals from wearable interface for augmenting the accuracy of the child monitoring system. The Fuzzy ARTMAP is an ART network for the association of analogy pattern in supervised mode and is capable of overcoming the stability-Plasticity dilemma. In our experiments, the arrays of sensor signals extracted from the wearable interface during monitoring process from toddlers are classified using the feature signal pattern. The high accuracy obtained as classification percentages validates the suitability of our proposed Fuzzy ARTMAP classification for such critical real time system.

Stable and crit gesticulation recognition in children and pregnant women by Naïve Bayes classification

The remote healthcare monitoring on a care taking base involves many implicit observations between the subjects and the care takers. Any ignorance and negligence leads to unpleasant situations thereafter. A wearable attire system can precisely interpret the implicit communication of the state of the subject and pass it to the care takers or to an automated aid device. Casual and conventional movements of subjects during play and living condition can be used for the above purpose. The proposed system suggests a novel way of identifying safe and unsafe conditions of playing for the children where a rapid warning assistance is required. The Same in the case of the normal and contraction time identification of pregnant women. Naive Bayes classifier was applied on five different sets of combinations of features constructed by Fractal Dimension, Fast Fourier Transformation, Singular Value Decomposition and combinations of the three. The result shows the combinational features of FFT and SVD are more supportive in all three sets of experiments and better classified by Navie Bayes classifier than the other combination and individual features. The experimental results show a well-distinguished realization of different body movement activities using a wearable attire array medium and the interpretation results always show significant and identifiable thresholds.