Entity

Time filter

Source Type


Madheswaran M.,Center for Research in Image and Signal Processing
International Journal of Applied Engineering Research | Year: 2015

In this paper, a distance based cluster head selection algorithm (DBR-LEACH) is proposed for improving the network life time, which estimates the minimum energy required for a node to become a cluster head by considering its residual energy and the distance of the node frombase station or sink node. The objective of this modified cluster head selection algorithm is to reduce the frequency of re-clustering and to ensure that nodes that are selected close to the base station possess more energy than the nodes that are farther to the base station. The experiment result shows that the proposed algorithm outperforms LEACH and D-LEACH protocol in terms of energy consumption, first node die time, alive nodes and thereby improved network life time. © Research India Publications.


Sumathi S.,Center for Research in Image and Signal Processing | Beaulah H.L.,Salem College
International Journal of Applied Engineering Research | Year: 2014

This paper approaches an intellectual diagnosis system using rule based evaluvation of myocardial infraction in electrocardiogram (ECG) signals using wavelet transform. This method is mainly through the computation of an indicator related to the area covered by the ST-wave curve. The algorithm is healthy to acquisition noise, to wave form morphological variations and to baseline wandering. To find the main computation has been implemented to simple finite impulse response filter. The inclusion of rule based system in the complex investigating algorithms yields very interesting recognition and classification capabilities across a broad spectrum of biomedical engineering. The results give importance to that the proposed model illustrates potential advantage to identify the myocardial Infraction. The sensitivity 90.01% and positive predictivity of 95.19 is achieved. © Research India Publications.


Ellappan V.,Center for Research in Image and Signal Processing | Samson Ravindran R.,Mahendra Engineering Colleges
Journal of Medical Imaging and Health Informatics | Year: 2016

In this research, an optimal wavelet filter coefficient design based methodology is proposed for the compression of medical image. The method utilizes new wavelet filter whose coefficients are derived by wavelet based genetic-group search optimizer (GGSO). The optimal wavelet coefficient based compression methods minimize the compression distortion, while Huffman and SPIHT coding further increase the compression without any loss of relevant image information. Overall, the proposed approach consists of three stages namely, (i) Designing of optimal coefficients, (ii) compression and (iii) decompression with optimal and adaptive approach. At first, the input medical images are taken to the training stages and the optimal filter coefficients are designed by the objective of improving compression ratio and PSNR. A comparative study of the performance of different existing approaches and the proposed compression approach is made in terms of compression ratio (CR), peak signal to noise ratio (PSNR), Cross-correlation, average difference and Normalize absolute error. When compared, the proposed wavelet filter gives better compression ratio and also yields good fidelity parameters. We can see that the optimal wavelet filter design based proposed method has outperformed by achieving better compression ratio of 10.87 when compared with the existing methods. © Copyright 2016 American Scientific Publishers. All rights reserved.


Madheswaran M.,Center for Research in Image and Signal Processing | Anto Sahaya Dhas D.,Rajas Engineering College
Biomedical Research (India) | Year: 2015

An enhanced classification system for classification of brain tumor from MR images using association of kernels with support vector machine is developed and presented in this paper. Oriented Rician Noise Reduction Anisotropic Diffusion filter is used for image denoising. A modified fuzzy c-means algorithm termed as Penalized fuzzy c-means algorithm is used for image segmentation. The texture and Tamura features are extracted using GSDM and Tamura method. Genetic algorithm with Joint entropy is adopted for feature selection. The classification is done by support vector machine along with various kernels and the performance is validated. A classification accuracy of 98.83% is obtained using SVM with GRBF kernel. © 2015, Scientific Publishers of India. All rights reserved.


Madheswaran M.,Center for Research in Image and Signal Processing | Shanmugasundaram R.N.,Center for Research in Image and Signal Processing
Wireless Personal Communications | Year: 2016

A modified weight based cluster head selection algorithm with balanced partitioning (BP-DCA) is proposed for the wireless sensor network and presented in this paper. This approach considers the node’s residual energy, number of neighbour nodes, average distance between the nodes for selecting the best nodes as cluster heads and distance between the cluster heads for optimum distribution of cluster heads during cluster formation for the next round. The experiment results show the efficacy of proposed algorithm in terms of energy consumption and prolonged network lifetime because of reduced overhead in less frequent cluster head selection. © 2016 Springer Science+Business Media New York

Discover hidden collaborations