Parthasarathi De

Identification of Tigers for Census by the Method of Tiger Iris Pattern Matching and Recognition

The Iris pattern of any animal (including human being) is statistically unique and suitable for biometric measurements. The identity of the animal concerned can be determined and verified comparing the templates obtained with the present algorithm with that template stored in database which was formed on the basis of previous studies. In the present study, the method of circular Hough transform is used for segmentation of the tiger Iris and subsequently Daugman‘s rubber Sheet model is used for normalization of the segmented values. Pattern matching is achieved by calculating Hamming Distance where its degree is proportional to the closeness of matching. The closer matching between the stored and calculated pattern is found to lead towards better recognition of Irises and thereby the animal itself.

Performance analysis of network monitoring tool through automated software engineering approach

In this paper, we developed an application that can be used to monitor network packets, bandwidth, Protocol Packets from the network and easily can monitor the system networking performance. The system uses automated software engineering approach, especially unit test in Java. The system can be used for monitoring the TCP, UDP, ARP, packets, when the system is connected with the internet. Domain IP address can be detected and packet that is lost can also be measured. The packet is captured with JPcap and WinPcap tool, integrated in JVM. Testing of this software involves developing set of test cases that are similar to the real data that the NMT is intended to manipulate. This project is a prototype of software engineering methodology and software is successfully tested.

Modified Method of Texture Feature Extraction and Analysis using Daubechies Wavelet and SVM

Text Extraction is always a challenging problem in the field of research area. In this paper we developed an algorithm that can be used to extract text from images. We have studied texture feature analysis and also successfully retrieve the texture features embedded in an image. We also calculated number of characters in blurred images. Our algorithm is successfully implemented on many images that are collected from ICDAR Dataset 2011, ICDAR 2013, ICDAR 2003 and some images are collected from newspapers, and digital camera. We used different methods for texture feature extraction. Our algorithm gives more efficiency and extract texture features as compared with previous studies.