Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/541412
Title: | Human Activity Recognition in Thermal Imagery using Deep Learning Techniques |
Researcher: | Srihari, Pasala |
Guide(s): | Harikiran, J |
Keywords: | 3D-CNN human action recognition Thermal images |
University: | Vellore Institute of Technology (VIT-AP) |
Completed Date: | 2023 |
Abstract: | Recognizing human activity in a camera-based surveillance system is still a difficult is- newlinesue. When a human is unseen in an image taken in a dark area, it is difficult to identify the action. Thermal cameras have been utilized in earlier investigations to address this problem. Frequently, for heat transfer inference, Infrared thermal imaging is a technique of passive newlineimaging that captures the discharged radiation from an object to calculate the temperature of newlinethe surface. Infrared thermal imaging provides the ability to detect movement without the newlineshadows, glare, or lighting modifications affiliated with imaging digital visualization or active infrared challenges.The identification and analysis of human action recognition in thermal im- newlineages are discussed in this work using a hybrid approach. One of the tough topics of research is the recognition of human action in the supervision video presently. To classify using tra- ditional algorithms of image processing the human actions comprised of the same patterns sequence that is hard. Video analysis is a significant field of research that implies analytics to the camera. It screens the contents of the video and abstracts intelligent data from it. The newlinemajority of the tasks in this field are subjected to constructing the classifying techniques on complex properties that are handcrafted or modelling DL-based CNNs, which work on inputs that are raw and take out important data along with the video directly. In this Thesis, for the segmentation of human activities in video sequences, k-means clustering is used. To classify and detect various human activities like boxing, carrying, digging, robbing, etc. the hybrid newlinecombination of ResNet50 and 3D-CNN is utilized. The (Resnet-50) Pre-trained technique is utilized as a DL technique in this article. In order to capture the information of motion among the adjoining frames, the 3D-CNN extracts the features in the dimension of temporal together newlinewith the dimension of spatial. The performance measures are evaluated for various met |
Pagination: | x,103 |
URI: | http://hdl.handle.net/10603/541412 |
Appears in Departments: | Department of Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf.pdf | Attached File | 55.54 kB | Adobe PDF | View/Open |
05_content.pdf | 46.62 kB | Adobe PDF | View/Open | |
06_abstract.pdf | 69.32 kB | Adobe PDF | View/Open | |
09_chapter 1.pdf | 514.81 kB | Adobe PDF | View/Open | |
10_chapter2.pdf | 139.69 kB | Adobe PDF | View/Open | |
11_chapter3.pdf | 930.8 kB | Adobe PDF | View/Open | |
12_chapter4.pdf | 845.73 kB | Adobe PDF | View/Open | |
13 _ chapter 5.pdf | 576.32 kB | Adobe PDF | View/Open | |
15_references.pdf | 116.09 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 46.05 kB | Adobe PDF | View/Open | |
prelim pages.pdf | 243.7 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: