Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/253360
Title: | Efficient object detection and movement tracking in h 264 compressed video using fuzzy sets |
Researcher: | Srinivasan K |
Guide(s): | Balamurugan P |
Keywords: | Engineering and Technology,Computer Science,Computer Science Information Systems fuzzy sets tracking |
University: | Anna University |
Completed Date: | 2018 |
Abstract: | Object detection is meant to detect the specific location and size newlineof a particular object in an image or a video scene. Object tracking is a newlinesignificant technique in the field of computer vision. With the growing need newlineof detection-based security and industrial applications, the object detection newlineand tracking in a fast and reliable manner has been attracting much interest. newlineIn this research, efficient object detection and movement tracking in h.264 newlinecompressed video is proposed. The first part of the work deals about newlinedetection and tracking with the help of fuzzy based optimal particle filter. newlineHere adaptive median filter is employed for preprocessing and the newlinemorphological operation is employed for segmentation. Finally object newlinedetection and object movement is selected by means of fuzzy and particle newlinefilters. Second part of the work is deals with similar object detection and newlinetracking in h.264 compressed video using modified local self similarity newlinedescriptor and particle filtering. Here foreground and background images are newlineseparated and then segmentation of object is carried out by using newlinemorphological operation. For similar object detection, the recommended newlinetechnique uses the modified local self-similarity descriptor and similar object newlinetracking is done by a particle filter. The performance of the proposed method newlinewas measured using evaluation metrics such as precision, recall, F-measure, newlineFPR, FNR, PWC, FAR, similarity, specificity, accuracy, FMR, FNMR and newlineGAR. Our proposed method is worked out with six different datasets of newlinemoving objects. newline newline |
Pagination: | xxiii, 160p. |
URI: | http://hdl.handle.net/10603/253360 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 17.52 kB | Adobe PDF | View/Open |
02_certificates.pdf | 1.34 MB | Adobe PDF | View/Open | |
03_abstract.pdf | 106.02 kB | Adobe PDF | View/Open | |
04_acknowledgment.pdf | 80.91 kB | Adobe PDF | View/Open | |
05_contents.pdf | 5.07 MB | Adobe PDF | View/Open | |
06_chapter1.pdf | 1.32 MB | Adobe PDF | View/Open | |
07_chapter2.pdf | 464.77 kB | Adobe PDF | View/Open | |
08_chapter3.pdf | 293.74 kB | Adobe PDF | View/Open | |
09_chapter4.pdf | 583.06 kB | Adobe PDF | View/Open | |
10_chapter5.pdf | 527.82 kB | Adobe PDF | View/Open | |
11_chapter6.pdf | 1.48 MB | Adobe PDF | View/Open | |
12_conclusion.pdf | 183.34 kB | Adobe PDF | View/Open | |
13_references.pdf | 276.78 kB | Adobe PDF | View/Open | |
14_publications.pdf | 989.26 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: