Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/302083
Title: | Automated detection and classification strategies for identification of marine animals from under sea water videos |
Researcher: | Jayachandra N. |
Guide(s): | Nadira Banu Kamal A.R. |
Keywords: | Computer Science Computer Science Information Systems Engineering and Technology |
University: | Mother Teresa Womens University |
Completed Date: | 2019 |
Abstract: | The development of automated recognition and classification system for fish varieties in the under sea water is essential for the fishers and marine managers to make accurate decisions regarding the by-catch fishing, population assessment and monitoring the ecosystems. The digital technologies are improving to provide long time period and good resolution underwater videos to identify objects using computer vision techniques. The main objective of this research work is to implement a machine vision system which results in identification of different types of sea animals present in under sea water videos. The image sets are created for all frames in these videos with different resolution. It helps to generate the best training set for each sea animal. newlineThe proposed methodology incorporates automated fragmentation of frames and construction of image set for all sea animals in a given video to identify the type of moving sea animals. For all different types of videos a common framework is given to detect the boundary region of moving objects in rough background by developing computer vision algorithm with the help of Optical Flow Detection. newlineThe proposed methodology helps to identify any type of video frames given as input without ground truth value to identify the type of sea animals. Moreover this methodology helps marine engineers to identify the rare sea animals and help the fishermen by indicating the living area of rare sea animals to provide the solution for bycatch problem. This achieves high accuracy in marine species detection and identification on both public and self-collected dataset with high uncertainty and class imbalance. Due to the basic significance of fish species recognition, numerous computer vision-based techniques are projected. These techniques can be mostly characterized into three main application areas based on their scope: Recognize and categorize the rare fish or sea animals in the controlled under sea water regions. |
Pagination: | v, 225p. |
URI: | http://hdl.handle.net/10603/302083 |
Appears in Departments: | Department of Computer Science |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 270.37 kB | Adobe PDF | View/Open |
02_certificate.pdf | 724.42 kB | Adobe PDF | View/Open | |
03_abstract.pdf | 95.82 kB | Adobe PDF | View/Open | |
04_declaration.pdf | 515.41 kB | Adobe PDF | View/Open | |
05_plagiarism certificate.pdf | 418.02 kB | Adobe PDF | View/Open | |
06_acknowledgement.pdf | 79.55 kB | Adobe PDF | View/Open | |
07_contents.pdf | 81.21 kB | Adobe PDF | View/Open | |
08_list of tables.pdf | 86.19 kB | Adobe PDF | View/Open | |
09_list of figures.pdf | 100.14 kB | Adobe PDF | View/Open | |
10_abbreviations.pdf | 89.79 kB | Adobe PDF | View/Open | |
11_chapter 1.pdf | 1.17 MB | Adobe PDF | View/Open | |
12_chapter 2.pdf | 271.02 kB | Adobe PDF | View/Open | |
13_chapter 3.pdf | 2.17 MB | Adobe PDF | View/Open | |
14_chapter 4.pdf | 2.03 MB | Adobe PDF | View/Open | |
15_chapter 5.pdf | 1.6 MB | Adobe PDF | View/Open | |
16_chapter 6.pdf | 1.26 MB | Adobe PDF | View/Open | |
17_conclusion.pdf | 109.93 kB | Adobe PDF | View/Open | |
18_bibliography.pdf | 161.96 kB | Adobe PDF | View/Open | |
19_appendix.pdf | 2.35 MB | Adobe PDF | View/Open | |
80_recommendation.pdf | 143.66 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: