Implementation and Analysis of Real-time Object Tracking on the Starburst MPSoC

Hakim, V.S. El (2015) Implementation and Analysis of Real-time Object Tracking on the Starburst MPSoC.

Abstract:Computer vision has experienced many advances over the recent years. As such, it started to rapidly find its way into many commercial and industrial applications, with the disciplines of visual object detection and tracking being most prominent. A good example of computer vision being applied in practice, is the social network and website Facebook, where advanced object detection algorithms are utilized to detect and recognize people and in particular -- their faces. One of the main reasons for the rapid expansion of computer vision in practice, is the ever emerging new computing platforms, such as the cloud service, capable to handle the complex mathematics involved in analyzing and extracting information from images. Despite its advances however, computer vision is still in its early stages of establishing a solid foothold in the world of embedded systems. Specifically, systems which are subjected to real-time constraints, with restricted computational resources, are struggling the most. Thus the usage of modern visual object detection and tracking algorithms in safety critical embedded systems is still more or less restricted. Fortunately, this is starting to change with newly developed embedded computer architectures, which employ application specific hardware to perform the task of computer vision more efficiently. In this thesis, two state-of-the-art computer vision algorithms in the form of HOG-SVM detection and Particle filter tracking, are explored and evaluated on a real-time embedded MPSoC, called Starburst. Eventually, it can be shown that these two seemingly ``difficult'' algorithms, not only can satisfy certain real-time constraints, but also achieve a high throughput on an embedded system such as Starburst. To accomplish this, the thesis contributes with a powerful real-time hardware architecture of the HOG object detector, and a software based multiple processor object tracking framework, based on the Particle filter. Both implementations are evaluated and tested on Starburst, to determine their respective real-time capabilities and whether imposed throughput constraints can be satisfied. Additionally, the functional behavior and accuracy of the implementations is also analyzed, but not to a full extent, since both algorithms are widely studied and documented in modern literature. The focus of this research is thus mainly on the temporal behavior.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics, 50 technical science in general, 53 electrotechnology, 54 computer science
Programme:Embedded Systems MSc (60331)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page