Object segmentation and visual tracking in digital marketing
| dc.contributor.author | Bohendiran A/L Balakrishnan | |
| dc.date.accessioned | 2021-03-01T02:56:04Z | |
| dc.date.available | 2021-03-01T02:56:04Z | |
| dc.date.issued | 2019-06 | |
| dc.description.abstract | This thesis describes the use of object segmentation and visual tracking tools to detect and track people in video sequences for digital marketing purposes. Closed Circuit Television (CCTV) footages were used to extract more beneficial information. Extracted information include customers counting and peak hour determinations. This thesis presents a method using an overhead camera in a zenithal position, to detect and count the number of customers entering or exiting a premise. Reduced customers’ privacy invasion is a key concern besides efforts to reduce occlusion problems. Methods of background subtraction, foreground detection, morphological filters, tracking algorithms and counting algorithms are experimented in this research. The aim of this research is to design a modular based people tracking and counting algorithm. The designed algorithm is able to substitute different image segmentation and visual tracking techniques to give output of high accuracy and be able to work in various environments. Based on five sample videos, a general range of parameters and combination of modules will be achieved and further tested and tuned on an experimental CCTV footage collected for this research. | en_US |
| dc.identifier.uri | http://hdl.handle.net/123456789/11608 | |
| dc.language.iso | en | en_US |
| dc.title | Object segmentation and visual tracking in digital marketing | en_US |
| dc.type | Other | en_US |
Files
License bundle
1 - 1 of 1
Loading...
- Name:
- license.txt
- Size:
- 1.71 KB
- Format:
- Item-specific license agreed upon to submission
- Description: