SCAD Institute of Technology

Palladam, India

SCAD Institute of Technology

Palladam, India
SEARCH FILTERS
Time filter
Source Type

Rajamanikkam N.,Arulmigu Palani Andavar College Of Arts | Ravichandran C.G.,SCAD Institute of Technology
Journal of Computational and Theoretical Nanoscience | Year: 2016

For efficient disease detection and treatment, images from different modalities are combined with help of fusion techniques. According to, currently, medical image fusion scheme is a challenging task in healthcare applications like diagnosis, research and treatment etc. In this study, a multimodality medical image fusion scheme is proposed and the performance is analyzed using quantitative measure. At first, two different pairs of image are taken as input: Magnetic Resonance Imaging (MRI) and Computed Tomography (CT). These pair of images is fused together using fusion formula, which formulated based on transparency-labeling to reduce the computational load. In addition, we state the problem as a discrete multi-label optimization using firefly algorithm, which combines squared error and weighted entropy. The proposed medical image fusion guarantees nearly global solutions, avoiding the pix elation artifacts and too much computation load. The resultant is analyzed using quantitative matrices like as mean square error (MSE). From the experimental results, it is observe that the proposed medical image fusion scheme is efficient in better quality of images which is inferred from the values of MSE. © 2016 American Scientific Publishers All rights reserved.


Gopalakrishnan R.C.,SCAD Institute of Technology
International Journal of Applied Engineering Research | Year: 2015

Motion can defined as to detect through measure change in speed or vector of an object or objects. Motion detection is extensively used in various ways in video surveillance application and to detect and track the human and human activities in real time video sequence. This paper is related to the extensive subject of motion detection and analysis in video surveillance of image sequence. Now a day in video surveillance application is used to detect multiple objects and monitor their activities are challenging task in indoor and outdoor environment. Consider spatio-temporal relationship among feature points, thereby enabling detection and classification of simple and complex human activities. In presence of a good number of real time problems such as, the problems are namely illumination changes, moving background and shadow detection. The robust algorithm is proposed for enhancing the accuracy and reliability of motion detection and classification methods to develop the real time video sequence in video surveillance. Its advantages are discussed and compared to the relative approaches for action recognition. The widely-used KTH human activity dataset demonstrated and they are the implemented state-of-the art methods. © Research India Publications.


Gopalakrishnan R.C.,SCAD Institute of Technology | Karunakaran M.,PSNA College of Engineering and Technology
PLoS ONE | Year: 2014

Nowadays, quality of service (QoS) is very popular in various research areas like distributed systems, multimedia real-time applications and networking. The requirements of these systems are to satisfy reliability, uptime, security constraints and throughput as well as application specific requirements. The real-time multimedia applications are commonly distributed over the network and meet various time constraints across networks without creating any intervention over control flows. In particular, video compressors make variable bit-rate streams that mismatch the constant-bit-rate channels typically provided by classical real-time protocols, severely reducing the efficiency of network utilization. Thus, it is necessary to enlarge the communication bandwidth to transfer the compressed multimedia streams using Flexible Time Triggered-Enhanced Switched Ethernet (FTT-ESE) protocol. FTT-ESE provides automation to calculate the compression level and change the bandwidth of the stream. This paper focuses on low-latency multimedia transmission over Ethernet with dynamic quality-of-service (QoS) management. This proposed framework deals with a dynamic QoS for multimedia transmission over Ethernet with FTT-ESE protocol. This paper also presents distinct QoS metrics based both on the image quality and network features. Some experiments with recorded and live video streams show the advantages of the proposed framework. To validate the solution we have designed and implemented a simulator based on the Matlab/Simulink, which is a tool to evaluate different network architecture using Simulink blocks. © 2014 Karunakaran, Gopalakrishnan.


Benadict Raja J.,PSNA College of Engineering and Technology | Ravichandran C.G.,SCAD Institute of Technology
International Journal of Engineering and Technology | Year: 2014

Diabetic macular edema is one of the retinal abnormalities which affects the central vision of the person and causes total blindness in severe cases. Fovea (center of macula) localization is an important step in retinal image analysis especially for grading diabetic macular edema. This paper describes a method to automatically localize the fovea center in retinal fundus images. The method is mainly based on mathematical morphology along with the information of other anatomic structures such as blood vessel and optic disc. Initially, the vascular structure and optic disc center are extracted, and then the morphological operations are employed on the gray scale image of green channel for fovea candidates' selection. The candidates' satisfying area, density and distance criteria are considered for the final stage. In the final stage, the candidate having lesser vessel pixels was considered as fovea region. The proposed method was evaluated on the two publicly available DRIVE and STARE databases. The method was able to obtain 100% of fovea localization accuracy on DRIVE database with 2.88 seconds average computation time.


Ravichandran C.G.,RVS College of Engineering and Technology | Siva Prakash P.,SCAD Institute of Technology
Asian Journal of Information Technology | Year: 2016

In today's digital age, law enforcement officials and even employers may find it easier than ever to take advantage of camera and wiretap surveillance. Surveillance cameras now line many public streets and workplace locations in an attempt to monitor activity and law enforcement agencies continue to use wiretapping to aid in investigations. Even with the advancement of technology, we have always resorted to manned surveillance techniques which will require impeccable human attention to the video feed received from the surveillance cameras. The orthodoxical way of surveillance system can be automated by spatially identifying the human body and the vital body parts namely head, torso, etc. from live video frames such as CCT V camera footage. With this spatial information from the video frames, we estimate the poses held with the temporal association between the successive and the previous video frames. The system works well with unconstrained backgrounds and without any premeditation of the clothing, brightness of the video frame. We also do not impose any constraints on the position of a person in the video frame. The only constraint that is imposed by our system is that people should be in a head-over-torso position with either near-frontal or near-rear viewpoint for greater accuracy of the estimation However, the system responded with a considerable accuracy for side poses as well, during testing. The poses gleaned are used for detecting punching activity. © Medwell Journals, 2016.


Ravichandran C.G.,SCAD Institute of Technology | Sivaprakash P.,RVS College of Engineering and Technology
International Journal of Control Theory and Applications | Year: 2016

In today's digital age, law enforcement officials and even employers may find it easier than ever to take advantage of camera and wiretap surveillance. Surveillance cameras now line many public streets and workplace locations in an attempt to monitor activity and law enforcement agencies continue to use wiretapping to aid in investigations. Even with the advancement of technology, we have always resorted to manned surveillance techniques, which will require impeccable human attention to the video feed received from the surveillance cameras. The orthodox call way of surveillance system can be automated by spatially identifying the human body and the vital body parts namely head, torso etc., from live video frames - such as CCTV camera footage. With this spatial information from the video frames, we try to roughly estimate the poses held with the temporal association between the successive and the previous video frames. One of the several challenges we face is the cluttered and natural background of the CCTV footages. Manned surveillance should be replaced only by a credible and reliable system. The system should be able to work with unconstrained backgrounds and without any premeditation of the clothing, brightness of the video frame. We also do not impose any constraints on the position of a person in the video frame. The only constraint that is imposed by our system is that people should be in a head-over-torso position with either nearfrontal or near-rear viewpoints. © International Science Press.


Ravichandran C.G.,SCAD Institute of Technology | Selvakumar R.,Madurai Institute Of Engg And Technology
Applied Computational Electromagnetics Society Journal | Year: 2016

Image fusion is place as a key role in medical image investigation and preparation of treatments for bio-medical research and clinical diagnosis. The most incentive is fuse to capture a large amount of vital information from the input images to have its output image. In this paper, a well-organized multimodal medical image fusion approach is obtainable to fuse computer tomography (CT) and magnetic resonance image (MRI). The significant co-efficient of source images are line up through the dual-tree complex wavelet transform (DTCWT), followed by unit of low and high frequency components. Two completely different proposed fusion rules based on weighted fusion rule, the weights are optimized by modified lion optimization technique (mLOT) and intensity co-variance verification (ICV) are used to fuse the low and high frequency coefficient. The fused image is reconstructed by inverse DTCWT with all amalgamate co-efficient. To prove the potency of the new approach is greater than the wellknown standard algorithm, experiments are conducted. Based on experimental comparison and proposed approach, the better results are fused image quality are obtained. The studies of qualitative and quantitative metrics are clearly demonstrated that the new approach is to display the high superior than the present. © 2016 ACES.


Kuppusamy V.,Bharath Niketan Engineering CollegeTamil Nadu | Gopalakrishnan R.C.,SCAD Institute of Technology
International Journal of Applied Engineering Research | Year: 2016

Lung cancer diagnosis using pattern classification is the active research topics in medical image processing. Feature extraction is considered as an essential step in medical image analysis and classification. Computer aided segmentation for computed tomography (CT) and magnetic resonance imaging (MRI) are finding the application in computer aided diagnosis, clinical studies, and treatment planning. In medical images typically suffer from one or more imperfections such as low resolution (in the spatial and spectral domains) and low contrast images. This work proposed the black circular neighborhood algorithm for feature extraction and genetic algorithm (GA) based clustering using the extracted nodules. The performance of our algorithm is to reduce number of false positives (FP) and results are improves the accuracy. © Research India Publications.


Ravichandran C.G.,Scad Institute of Technology | Venkateshbabu S.,Anna University
Journal of Medical Imaging and Health Informatics | Year: 2016

In general medical image processing can get benefit by computerized tomography from hardware acceleration. To design this intensive medical image processing algorithm, requires the rapid processing of large data. In this paper, we proposed Error-Correction code to remove the transient error and permanent error for combinational circuits. The error-correction is done by implementing Error-correcting code, by storing the adder output in errortolerant memory elements. The duplicated adder circuit of sum-bit will be monitored by ECC circuit based on the ECC bit generator. The error is detected by checking the adder output from memory and compared with the ECC code. And error-correction code corrects the error by sending the error-detected signal to the register. And the correct output values are analyzed by comparing the output in the duration of transient error from previous output. After recovering the error, the system will continue to work in the error-corrected state. The result is the complete design of error-correction adder, capable for efficient solutions for self-correcting architectures. This proposed implementation, can easily adapted to different range of tomographic scanners. Copyright © 2016 American Scientific Publishers All rights reserved.


Sathish Kumar N.,SCAD Institute of Technology | Ravichandran C.G.,SCAD Institute of Technology
Asian Journal of Information Technology | Year: 2016

In the early 1980's evolution of wireless communication has witnessed remarkable growth in increasing data rates ever since its evolution. Long Term Evolution (LTE) is shaping as the future of next generation technology to satisfy the demand for growing data rates. An enhanced scheduling scheme is very essential in achieving an effective Quality of Service (QoS) Provisioning in LTE systems because it plays a significant role in distributing the wireless resources of the system among mobile users effectively. It also managing and guarantying the QoS requirements of the network responsibly. This research offers more flexibility in developing the network to design an appropriate scheduling algorithm. This study proposed a new dimension of packet scheduling scheme to made scheduling decisions using the channel quality conditions without ignoring the fairness criteria. Unlike the traditional schemes this optimum scheduling scheme for LTE system also supports multiple users with different QoS requirements within same class. The same QoS requirements for different users within each class are assumed in this scheme. The scheme also ensures maintaining a high QoS with a good throughput-fairness trade-off ratio. © Medwell Journals, 2016.

Loading SCAD Institute of Technology collaborators
Loading SCAD Institute of Technology collaborators