An intergration of thermal and optical flow technique for human aggressive movement detection
There is growing interest in intelligent video surveillance as for public security has become more and more important especially after the attack of 11 September. The goal in developing intelligent video surveillance is to replace the traditional passive video surveillance that is proved to be ineff...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://ir.upnm.edu.my/id/eprint/31/1/AN%20INTERGRATION%20OF%20THERMAL%20%28Full%29.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | There is growing interest in intelligent video surveillance as for public security has become more and more important especially after the attack of 11 September. The goal in developing intelligent video surveillance is to replace the traditional passive video surveillance that is proved to be ineffective as the number of cameras exceeds the capability of human operators to monitor them. In this real world, being able to identify the signs of imminent aggressive behaviours such as aggression or violence and also fights, is of extreme importance in keeping safe those in harm’s way. This research proposes an approach to figure out human aggressive movements using two methods which are movement based and colour based. For the movement based, Horn-Schunck
optical flow algorithm is chosen in order to find the flow vector for all video frames. Optical flow is a popular method to detect the object and can calculate the motion of
each pixel between two frames, and thus it provides a possible way to get a velocity of the object movement. The video frames are collected using the digital camera and
thermal camera. This research guides and discovers the patterns of body distracted movement so that suspect of aggression can be detected automatically without body
contact. Using this method, the aggressive and non aggressive video frames are then analysed and utilised to define the aggressiveness of humans. This research embarks on the following objectives which are to extract the suitable features that can represent aggressiveness, to develop an algorithm that can discriminate between aggressive and non - aggressive features for human movement detection and to adapt into the digital and thermal video images. The experiments conducted carried out to compare the HornSchunck algorithm under different types of images. Overall the combination of thermal images and Horn-Schunck optical flow proves to be able to accurately distinguish,
detect and track the human aggressiveness. In future work, the intelligence system for human movement detection that can be applied at the Malaysia border area as a virtual
guard system. |
|---|