| Author | Thiparat Jirawattanakitja |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Computer Science, School of Engineering and Technology
|
| Publisher | Asian Institute of Technology |
| Abstract | Object detection has applications in many areas of computer vision, including image retrieval and video surveillance. Face and license plate detection are especially key steps in many processing systems nowadays. In this thesis, I develop a system to collect information about cars’ and drivers’ identities in a parking area in order to improve security of parking lots. We mainly experiment with data capture from guardhouses at the entrances and exits of parking areas. In case a crime such as license plate switching or car theft occurs, the system can help provide evidence to point out who the thief is. In my detection experiments, I choose two interesting approaches. The first algorithm is the Adaboost (Adaptive Boosting cascade) training algorithm. This method is very effective and widely used for object detection. The other algorithm is hyperfeatures. This method has very good results in recognition problems. I apply Adaboost cascading and hyper-features to the two object detection problems. I summarize the overall performance and error patterns of each method in terms of ROC analysis for comparison.
|
| Year | 2012 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Dailey, Matthew N.; |
| Examination Committee(s) | Honda, Kiyoshi;Mongkol Ekpanyapong ; |