Robust Mean Shift Tracking Based on Refined Appearance Model and Online Update
In this paper, a robust mean shift tracking algorithm based on refined appearance model and online update strategy is proposed. The main idea of the proposed algorithm is to construct a more accurate appearance model and design an online update strategy.
- PDF / 1,097,592 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 65 Downloads / 200 Views
Abstract. In this paper, a robust mean shift tracking algorithm based on refined appearance model and online update strategy is proposed. The main idea of the proposed algorithm is to construct a more accurate appearance model and design an online update strategy. At the beginning of the tracking, the simple mean shift tracking algorithm is applied on the first few frames to collect a set of target templates, which contains both foreground and background of the target. During the model construction, simple linear iterative clustering (SLIC) algorithm is exploited to obtain the superpixels of the target templates, and the superpixels are further clustered to classify the background from foreground. A weighted vector is then obtained based on the classified background and foreground, which is utilized to modify the kernel histogram appearance model. The following frames are processed based on the mean shift tracking algorithm with the modified appearance model, and the stable tracking results with no occlusion will be selected to update the appearance model. The concrete operation of model update is the same as model construction. Experiment results on challenging test sequences indicate that the proposed algorithm can well cope with both appearance variation and background change to obtain a robust tracking performance.
1
Introduction
With the development of computer vision and multimedia technology, visual tracking has been widely applied in many civil and military fields. It plays more and more important roles in improving the efficiency of industrial and agricultural production, as well as the performances of weapons and equipment. In the past decade, visual tracking technology made much progress [1]. However, there exists limitation in most of the tracking methods because they are designed for the specific or relatively simple situations [2]. As one of the famous tracking methods, mean shift tracking attracts many attentions for the well-developed theory, simple course, outperformed performance and easy to implement. The mean shift algorithm was firstly proposed by Fukunaga et al. [3] to cope with the data analysis. Cheng [4] introduced it into the fields of image processing and computer vision. Bradski [5] This research was supported by National Natural Science Foundation of China (No. 61175029 and No. 61473309) c Springer-Verlag Berlin Heidelberg 2015 H. Zha et al. (Eds.): CCCV 2015, Part I, CCIS 546, pp. 114–123, 2015. DOI: 10.1007/978-3-662-48558-3 12
Robust Mean Shift Tracking Based on Refined Appearance Model
115
developed its application in face tracking and proposed a continuously adaptive mean shift (CAMSHIFT) algorithm. Comaniciu et al. summarized the mean shift as a robust approach to feature space analysis [6] and successfully applied it to visual tracking [7]. Collins [8] discussed the limitation of the scale adaptation in original mean shift tracking algorithm, and proposed a modified one in scale space. Zivkovic et al. [9] and Ning et al. [10] also discussed the scale and orientation estimation for mea
Data Loading...
