中文摘要：

为了研究无约束环境下长时间可视跟踪问题,提出了一种在线学习多重检测的对象跟踪方法.该方法以随机蕨作为基础检测器结构,通过在线学习的方式,将目标对象的整体和局部表观,以及由场景学习中发掘的同步对象同时作为检测学习的基础数据,该检测器因而具备了对这多种对象的独立检测能力.由于其各个检测部分发挥了各自不同的作用,本文从测量的角度将检测器对这三种对象检测的结果进行融合,通过计算检测关于目标的配置概率进而确定目标位置,实现对象跟踪任务.基于真实视频序列的实验结果验证了本文方法的有效性和稳定性,以及较现有的跟踪方法在跟踪性能上的提高.

英文摘要：

In order to study the problem of long-term visual tracking in unconstrained environments, this paper proposes a method of learning multiple detectors online for visual object tracking. The method uses the random ferns as the basic detector. The entire and the local appearances of the target and the connected objects which are explored by the context learning are used synchronously as the training data to build and upgrade the object detector on the fly. Thus it is able to detect the objects with different classes independently. Since different detections are related to different object classes, the results of object detections are fused as the measurements and the probabilities of configuration hypotheses for the measurements to the target are calculated to find the target location for visual tracking task. Experimental results based on the real-world video sequences validate the effectiveness and robusmess of our approach and demonstrate its better tracking performance than several state-of-the-art methods.