|
|
| |
| Articles
by
P. Balasubramanie |
Total Records (
4 ) for
P. Balasubramanie |
|
 |
| |
|
| |
B. Nagarajan
and
P. Balasubramanie
|
| |
Object classification in static images is a difficult task since motion information in no longer usable. The challenging task in object classification problem is the removal of cluttered background containing trees, road views, buildings and occlusions. The goal of this study is to build a system that detects and classifies the car objects amidst background clutter and mild occlusion. This study addresses the issues to classify objects of real-world images containing side views of cars with cluttered background with that of non-car images with natural scenes. The threshold technique with background subtraction is used to segment the background region to extract the object of interest. The background segmented image with region of interest is divided into equal sized blocks of sub-images. The statistical features are extracted from each sub-block. The features of the objects are fed to the back-propagation neural classifier. Thus, the performance of the neural classifier is compared with various categories of block size. Quantitative evaluation shows improved results of 83.8%. A critical evaluation of our approach under the proposed standards is presented. |
|
|
| |
|
| |
B. Nagarajan
and
P. Balasubramanie
|
| |
Object recognition and classification in a multi-environment is an important part of machine vision. The goal of this study, is to build a system that classifies the objects amidst background clutter and mild occlusion. This study addresses the issues to classify objects of real-world images containing side views of cars with cluttered background with that of non-car images with natural scenes. The threshold technique with background subtraction is used to segment the background region to extract the object of interest. The background segmented image with region of interest is divided into rectangular sub-images of equal size. The moment features which are invariant to Rotation, Scaling and Translation (RST) are extracted from each rectangular block. The features of the objects are fed to the back-propagation neural classifier. Thus, the performance of the neural classifier is compared with various categories of rectangular block size. Quantitative evaluation shows improved results of 84.9%. A critical evaluation of our approach under the proposed standards is presented. |
|
|
| |
|
| |
J. Premalatha
,
P. Balasubramanie
and
C. Venkatesh
|
| |
In mobile ad hoc wireless networks, multiple mobile stations communicate without the support of a centralized coordination station for the scheduling of transmissions. The dynamic nature of these mobile networks makes the support of quality of service challenging, because it is very difficult to maintain resource reservations. This study deals a combination of medium access control procedure employing distributed coordination function and suitable transport layer mechanism which improves QoS guarantee in Transport layer. IEEE 802.11e MAC employs a channel access function called the hybrid coordination function, which includes contention based channel access and a contention-free centrally controlled channel access mechanism. In the proposed method, IEEE802.11e and Additive Increase Multiplicative Decrease (AIMD) mechanism have been combined to analyze the quality of service in cross layer. The combined technique enhances the QoS parameters viz, throughput by 30-40% and decreases delay by 70-80% and packet loss by 30-40%. |
|
| |
|
| |
R.C. Suganthe
and
P. Balasubramanie
|
| |
A Mobile Ad Hoc Network (MANET) is a collection of mobile nodes that can communicate with each other using Multihop wireless links without using any fixed infrastructure and centralized controller. Communication links are susceptible to frequent failures due to intervening objects, which can cause intermittent connectivity i.e., there is no end-to-end path exists between source and destination all the time. Existing ad hoc routing protocols unable to deliver packets in the presence of a network partition between source and destination since they are robust to rapidly changing network topology. Flooding based schemes and Message Ferrying schemes are proposed by many researchers to overcome network partitions in intermittently connected ad hoc network. Flooding scheme is not suitable if partitions last for a long duration of time. Message Ferry distributes messages between nodes which are located in different partitions which may be disconnected. Ferry moves around a fixed path for providing regular connectivity in a disconnected network. But this scheme needs huge buffer space and also online collaboration between Ferry and other nodes in the network. With this in mind, a new routing scheme with two types of Ferries has been proposed. This scheme improves delivery rate and delay and it does not need any online collaboration between ferry and mobile nodes, it needs only online collaboration between local and global ferries. |
|
|
|
|
| |
|
