Asian Science Citation Index is committed to provide an authoritative, trusted and significant information by the coverage of the most important and influential journals to meet the needs of the global scientific community.  
ASCI Database
308-Lasani Town,
Sargodha Road,
Faisalabad, Pakistan
Fax: +92-41-8815544
Contact Via Web
Suggest a Journal
Articles by C.S. Ravichandran
Total Records ( 8 ) for C.S. Ravichandran
  C.S. Ravichandran , S. Subha Rani and Manigandan. T
  Modeling physical systems usually results in complex high-order dynamic models. It is often desirable to approximate these models by simpler models with reduced order. This study deals with the design of discrete time linear system using a balanced approach reduced order model. The reduced order model retains the desired state variable which contains a significant contribution. A PID controller is designed for the reduced second order model to meet the desired performance specifications by using pole-zero cancellation method. The stabilization of linear discrete time system is achieved by selection of parameters of the PID controller. A numerical example is given to illustrate the design method.
  M. Ravichandran and C.S. Ravichandran
  As network speed increases tremendously than CPU and memory speed, it widens the gap between network and the end system. This poses a major challenge in the field of intrusion detection as the data has to be scanned at link speeds to detect the malicious packets. The prominent solution for this problem is addressed by the invent of multicore processors which cope up with increasing link speeds by offering parallelism and the required instruction sets to provide the necessary throughput. By making use of the multicore Network Processors it is possible to parallelize the intrusion detection activity at the data rate. In this scenario where processor is subject to computationally intensive task, performing stateful signature based analysis together with dynamic load balancing, without efficient parallelization is thorniest. Hence, an enhanced way of parallelization technique is proposed in this study to dynamically distribute the load among the core without jeopardizing the analysis reliability.
  T. Vandarkuzhali and C.S. Ravichandran
  Fundus image analysis it is very difficult to identify the fovea region and eye diseases. Some times its cause to achieve a success to meet the cost and importance. An automated fundus image analysis system is developed for the detection of optic disc, blood vessels, fovea, etc. And also the identification of different eye diseases. The detection and analysis are proceeds in three different stages. They are extract the candidate region (preprocessing), features extraction and classification. Here, the optic disc localisation stage and a pre-processing stage to reduce noise and blood vessel structures and finally classify the fovea region. Also identifying the disease using the neural network classifier. For GF-SVM mechanisms. The optic disc localisation results in a localised point that represents the centroid of optic disc region whereas optic disc segmentation results in a complete contour of optic disc. In the optic disc localization stage, a feature vector approach that employs four salient characteristics of the optic disc is implemented. Fovea is one of the important feature of a fundus image. Fovea detection helps doctors and non-trained persons to identify Diabetic Retinopathy (DR), Age Related Macular Degeneration (AMD), Retinopathy of Pre-maturity (ROP) and some other diseases of the patients. Diabetic retinopathy is a cause of sight loss sometimes it will reach an advanced stage and cannot be cure. However, retinal image is essential and crucial for the ophthalmologists to diagnosis the disease. In the RGB image the green channel exhibits the best contrast between the vessels and background. With the help of advanced adaptive histogram equalization, thresholding method and smoothening method can detect the fovea region. Gabor filter and support vector machine are also used for classifying the features and its similar parts. The automatic screening will help for the doctors to quickly identify the condition of patients. Here, implemented a new efficient method to localize the fovea in retinal fundus image. Also, it is the new integrated efficient method to detect both disease and an eye region. In this proposed research aim for automatic screening of fovea for detection of many diseases quickly at a time. By automatically identifying the normal images, the workload and its costs will be reduced by increasing the effectiveness of the screening programs. The data base collected from Lotus Eye Hospital, Coimbatore. According to data, we can detect the sensitivity, specificity, accuracy, etc.
  T. Vandarkuzhali and C.S. Ravichandran
  Retinal image analysis is extremely important in medical image processing. Diabetic Retinopathy (DR) is an eye disease that can lead to complete loss of visual capacity, if left undiagnosed at the initial stage. In India DR is the 3rd cause of blindness. Diabetic retinopathy is obtained automatically. A computer has used to predict the qualitative research and emerging knowledge about retina using extreme learning machine classifier. The fundus image analysis developed to assist ophthalmologist’s diagnosis and also functions as an automatic tool for the mass screening of diabetic. In this method, extreme learning machine is used to detect the abnormal image. Texture features are extracted by using Gray Level Co-occurrence matrix (GLCM). Fovea is one of the important feature of a fundus retinal image. During the last 30 years, people are trying to extract the different features like blood vessels, optic disk, macula, fovea automatically from retinal image. The retinal image of a person, processing and pattern recognition can be performed. It can differentiate the bright region (optic disc) and dark region (fovea). While, compares the similarity or dissimilarity between regions can detect fovea region. The architecture gets the retinal image acquired from fundus camera and pre-process the image using histogram equalization, performs the segmentation algorithm for detecting the blood vessels, optic disk and fovea.
  K. Meenakshi Sundaram and C.S. Ravichandran
  Medical diagnostic and imaging system are ubiquitous in modern health care facilities. The advantages of early detection of potential lesions and suspicious masses within the bodily tissue have been well established. It can be detected and assessed many different types of injuries, diseases and conditions with the aid of the medical imaging that allows medical personnel to look into living cells non-instructively. Chronic Obstructive Pulmonary Disease (COPD) is the fourth leading cause of death worldwide and the only chronic disease with increasing mortality rates. COPD is the name for a group of lung diseases including chronic bronchitis, emphysema and chronic obstructive airways disease. This study involves in improving the accuracy over the existing technique using the adaptive region growing property and Adaptive-Neuro-Fuzzy Inference System-ANFIS classifier. Initially, pre-processing is carried out for the input image by Adaptive median Filter technique to make the image suitable for further processing. The contours of the image will be obtained using region growing technique. The ANFIS classifier is then used to confirm the suspected COPD cavities. The classification will be carried out by the features which have been taken from the segmented image. The proposed technique is implemented in MATLAB and the performance is compared with the existing technique. From the experimental result it can be said that the proposed method achieved more accuracy as compared with existing techniques.
  M. Belsam Jeba Ananth and C.S. RaviChandran
  This study presents a Particle Swarm Optimization (PSO) Method for determining the optimal Proportional-Integral Derivative (PID) controller parameters for speed control of a linear brushless DC motor. The proposed approach has superior features including easy implementation, stable convergence characteristic and good computational efficiency. The brushless DC motor is modeled in Simulink and the PSO algorithm is implemented in MATLAB. Comparing with Genetic Algorithm (GA) and Linear Quadratic Regulator (LQR) Method, the proposed method was more efficient in improving the step response characteristics such as reducing the steady-states error; rise time, settling time and maximum overshoot in speed control of a linear brushless DC motor.
  M. Ravichandran and C.S. Ravichandran
  An Anomaly based Intrusion Detection System is a one which monitors the system or network traffic looking for anomalous behaviour rather than matching the user behaviour pattern alone. Hence, the Anomaly Based Intrusion Detection algorithms have the capability to extend their detection mechanisms to detect unknown attacks. In this research, a Self Learning algorithm for anomaly based Intrusion Detection Model which is based on genetic neural network is proposed. The genetic neural network combines the good global searching ability of Genetic algorithm with the accurate local searching feature of back propagation neural networks. Here, it is used to optimize the initial weights of the neural network. The scope of the algorithm in this proposed research remains in identifying the malicious packet.
  M. Ravichandran and C.S. Ravichandran
  Intrusion Detection Systems (IDS) are progressively becoming a key part of network security by supervising the network traffic and attempting to identify and alert all malicious behavior to the user. In this study, Snort, a Signature-Based IDS that identifies attacks in a network traffic based on the rules written and SPADE, a Statistical-Based IDS that flags for anomalous behavior in the network traffic by learning the normal behavior for the packets in that particular network are both used. The integration of Snort and SPADE improves the detection accuracy and a vast majority of anomalous traffic on the network can be identified. However, the system is still prone to false positive errors that occur when a normal activity is misclassified as an attack. This leads to the anomaly based IDS to produce undesirable results as normal packets are being classified as malicious packets and are dropped or rejected by the system and also raising frequent false alerts. The main objective of this research is to find a scenario in which false positives are generated by SPADE and proceed to modify SPADE in such a way that it can be deployed effectively in a Wireless Ad Hoc Network.
Copyright   |   Desclaimer   |    Privacy Policy   |   Browsers   |   Accessibility