Visual tracking using convolutional features with sparse coding
- PDF / 1,805,068 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 88 Downloads / 205 Views
Visual tracking using convolutional features with sparse coding Mohammed Y. Abbass1,2 · Ki‑Chul Kwon1 · Nam Kim1 · Safey A. Abdelwahab2 · Fathi E. Abd El‑Samie3,5 · Ashraf A. M. Khalaf4
© Springer Nature B.V. 2020
Abstract Visual object tracking has become one of the most active research topics in computer vision, and it has been applied in several commercial applications. Several visual trackers have been presented in the last two decades. They target different tracking objectives. Object tracking from a real-time video is a challenging problem. Therefore, a robust tracker is required to consider many aspects of videos such as camera motion, occlusion, illumination effect, clutter, and similar appearance. In this paper, we propose an efficient object tracking algorithm that adaptively represents the object appearance using CNN-based features. A sparse measurement matrix is proposed to extract the compressed features for the appearance model without sacrificing the performance. We compress sample images of the foreground object and the background by the sparse matrix. When re-detection is needed, the tracking algorithm conducts an SVM classifier on the extracted features with online update in the compressed domain. A search strategy is proposed to reduce the computational burden in the detection step. Extensive simulations with a challenging video dataset demonstrate that the proposed tracking algorithm provides real-time tracking, while delivering substantially better tracking performance than those of the state-of-the-art techniques in terms of robustness, accuracy, and efficiency. Keywords Object tracking · Convolutional features · Compressive sensing · SVM
1 Introduction One of the major tasks in the field of computer vision is to enable machines such as robots, computers, drones, and vehicles to perform the main tasks of the human vision system like image comprehension and motion analysis (Abbass et al. 2018). To realize the function of intelligent motion analysis, many works have studied visual object tracking, which is becoming a highly-demanded research area in the real-time computer vision field (Abbass et al. 2019). The main step of visual tracking is to evaluate the trajectory model (i.e., position, direction, shape, etc.) of a tracked object in each scene of a video sequence (Wang * Nam Kim [email protected] Extended author information available on the last page of the article
13
Vol.:(0123456789)
M. Y. Abbass et al.
et al. 2016). A robust tracker appoints consistent markers to the target objects in successive scenes. In short, visual tracking is an operation that seeks to locate, detect, and define the dynamic positions of objects in the video sequence of one or multiple cameras (Alam et al. 2020; Li et al. 2017; Abbass et al. 2020a, b, c). Recently, Kernelized Correlation Filter (KCF) trackers have shown great efficiency. These KCF trackers use the circulant structure of the training features and implement the filters in the frequency domain to achieve high computational efficiency.
Data Loading...
