In human tracking, sparse representation successfully localises the human in a video with minimal reconstruction error using target templates. However, the state-of-the-art approaches use colour and local appearance o...
详细信息
In human tracking, sparse representation successfully localises the human in a video with minimal reconstruction error using target templates. However, the state-of-the-art approaches use colour and local appearance of a human to discriminate the human from the background regions, and hence fail when the human is occluded and appears in the varying illumination environment. In this study, a robust tracking algorithm is proposed that utilises gradient orientation and fine and coarse sparse representation of the target template. Sparse representation-based human appearance model utilises weighted gradient orientation that is insensitive to illumination variation. Coarse and fine representation of sparse code facilitates tracking under varying scales. Subspace learning from image gradient orientation is enforced with occlusion detection during the dictionary updation stage to capture the visual characteristics of the local human appearance that supports tracking under partial occlusion with lesser tracking error. The proposed human trackingalgorithm is evaluated on various datasets and shows efficient human tracking performance when compared to the other state-of-the-art approaches. Furthermore, the proposed human trackingalgorithm is suitable for surveillance applications.
As a classic appearance modelling method in object tracking, patch-based approach is believed to own natural superiority in handling local occlusion to its divide-and-conquer philosophy. However, in facing of more sev...
详细信息
As a classic appearance modelling method in object tracking, patch-based approach is believed to own natural superiority in handling local occlusion to its divide-and-conquer philosophy. However, in facing of more severe application conditions, such as heavy occlusion, part deformation and illumination change, traditional patch-based method may also fail due to the lack of sufficient matching patches. To address this problem, temporal stability as well as spatial salience to collaboratively improve patch selection and update schemes, resulting in a robust tracking algorithm for more challenging scenarios are proposed. Both quantitative and qualitative experiments conducted on practical video sequences demonstrate the effectiveness of the proposed method.
暂无评论