Player detection using one-class SVM
Xuefeng Bai, Tiejun Zhang, Chuanjun Wang, Qiong Li, Xiamu Niu · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2012
In this paper, a novel player detection method via One-Class SVM(OCSVM) is proposed, inspired by both the player detection problem and the property of the OCSVM. In this detection method, candidate regions are got by local entropy and local range analysis firstly. Then a set of training samples is obtained by several predefined rules on shape and area. These samples are used to train two OCSVM models. One model uses color feature, and the other uses gradient feature. Finally, we locate the regions of player by fusing the detection result of the two models. Extensive experiments demonstrate effectiveness and efficiency of the proposed method.