Automatic Track Creation and Deletion Framework for Face Tracking

摘要

The proposed approach consists of improving the track management by the creation and deletion of the track when occlusion or failure occurs. In this approach multiface tracking can be possible. Track creation and deletion will avoid errorness failure and improve track management. We improve the accuracy of face detection by using cascade classifiers. Also the face tracking is improved by using Haar Cascade algorithm. Haar cascade, very rarely addressed in the literature, is difficult due to object detector deficiencies or observation models that are insufficient to describe the full variability of tracked objects and deliver reliable likelihood (tracking) information. To achieve this, long-term observations from the image and the tracker itself are collected and processed in a principled way using decision tree algorithm, deciding on when to add and remove a target to the tracker. Proposed algorithm increases the performance considerably with respect to state-of-the-art tracking methods not using long-term observations and HMMs.