Automatic Broadcast Soccer Video Analysis, Player Detection, and Tracking Based on Color Histogram
In this chapter, a broadcast soccer video analysis system is proposed for the detection and tracking of the players. Our method consists of two phases. The first one is the scene analysis phase which automatically classifies the video into different scene
- PDF / 290,775 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 31 Downloads / 207 Views
Abstract In this chapter, a broadcast soccer video analysis system is proposed for the detection and tracking of the players. Our method consists of two phases. The first one is the scene analysis phase which automatically classifies the video into different scenes based on 2-D Gaussian color model of hue and saturation. An adaptive incremental model update scheme is proposed so that even under shadow condition, good scene analysis result can still be provided. The second one is the player analysis phase. A color histogram-based method is proposed to classify the player with a decision tree, and a linear prediction model based on spatial similarity matrix (SSM) is used for tracking of the players. Experimental results show that the proposed method is simple yet effective. Keywords Color histogram • Playfield segmentation • Player detection/tracking • Adaptive Gaussian color model update • Spatial similarity matrix (SSM)
1 Introduction Automatic soccer video content analysis has become an extremely interesting research topic. Among the various features used in soccer scene classification, color information is most useful. The playfield color distribution and playfield ratio are often used for soccer video shot classification [1–3]. The color distance between a pixel and the mean value of the playfield is used to determine whether it belongs to the playfield or not [4]. Different color models and different
D.-J. Duh (*) Department of Computer Science and Information Engineering, Chien Hsin University of Science and Technology, Tao-Yuan 32097, Taiwan e-mail: [email protected] S.-Y. Chang • S.-Y. Chen • C.-C. Kan Department of Computer Engineering and Science, Yuan Ze University, Chung-Li 32003, Taiwan J. Juang and Y.-C. Huang (eds.), Intelligent Technologies and Engineering Systems, Lecture Notes in Electrical Engineering 234, DOI 10.1007/978-1-4614-6747-2_15, # Springer Science+Business Media New York 2013
123
124
D.-J. Duh et al.
Initial Processing & Boundary Detection
Segmentation
Scene Analysis
Player Detection
Player Tracking
Fig. 1 The proposed system function blocks
segmentation scheme are used, such as simple thresholding method that chooses the color range for thresholding [5], color distribution of probability and statistics, and other solutions that use color histograms. Generally, the image is divided into several blocks, and the ratio between the blocks can be used to classify the scene. After the playfield segmentation, player detection and tracking process is used. Most of the algorithms for moving object detection are mainly based on background subtraction. Features, such as edge, skeleton, shape, Haar features, and SVM classifier, are used to identify the true players. Multiple player candidates can be detected after the player detection in each frame. Many existing algorithms for the multiple target tracking in different frames are used, such as Kalman filter and Bayesian tracker. Generally, when the playfield contains shadow, the results of segmentation and player detection are not good
Data Loading...
