Secondary:
School - Collegiate School, Krishnanagar, West Bengal, India
Passing Year - 2004
Aggregate - 79.97%
Higher Secondary:
School - Collegiate School, Krishnanagar, West Bengal, India
Passing Year - 2006
Aggregate - 82.10%
Prediction and generation of states for specific inputs in Null and Periodic boundary cellular automata.
Thesis Title: Archival and Retrieval of Heritage Images
Work: In this work, a study of a knowledge and information system to archive diverse Indian cultural heritage information has been made. A community based knowledge-sharing web-based system namely, SMart Archival Retrieval And Knowledge sharing (SMARAK), has been designed to archive heritage images using such knowledge representation. The information which are gathered from different users, may lack authenticity and may be incomplete. A method to dynamically annotate incomplete information of an image using a crowd-sourcing strategy is proposed in this work. To avoid redundancy in storage due to multiple instances of the same images, a technique for detection of duplicate images is developed. A domain specific vocabulary is built using related historical terms which is used to search a string in textual information of heritage images. Using the inter-relationship of different historical entities, structured text based queries are also supported. Content based image search is performed by local feature based and bag of visual word based searching methods. However, both the methods have a large response time in retrieving heritage images of high resolution. In this work, parallel search techniques using message passing interface (MPI) have been used to reduce the response time in content based queries.
Thesis Title: Land Cover Analysis Of Coal Surface Mine Regions Using Multi-Spectral Images.
Work: Surface mining is a widely used technique, where a shallow ore deposit is accessed by removing layers of soil or rock above it, which hosts it. It causes huge land use and land cover changes and directly impacts the eco-environment, soil fertility, water pollution, regional biodiversity, etc. Surface coal mining has additional adversity of coal seam fire, which emits poisonous gases. Classification and land cover monitoring of surface coal mine regions have several research challenges. Mostly, supervised and semi-supervised classification techniques are used in literature. Motivation of this thesis is to detect several land classes of surface coal mine regions, such as, mine quarry, mine dump, mine water body, overburden dump regions, coal seam fire locations, by exploiting their spectral characteristics in an unsupervised manner. Spectral ratio of short wave infra-red one and short wave infra-red two has been found instrumental to detect these regions in adaptive fashion. A novel index has been proposed to detect coal mine regions. Further, connected component analysis and hierarchical k-mean clustering have been used to detect other coal surface mine land classes such as mine water bodies, coal overburden dump, etc.
2024 (Feb) - 2025 (Feb)
University of Pavia, Italy (QS - 469).
Supervisor: Prof. Fabio Dell'Acqua
Work - Inter-row Management of Vineyards through Sentinel-2