It has been shown that the dynamic environment around the mobile robot can be efficiently and robustly represented by the Bayesian occupancy filter(BOF). In the BOF framework, the physical world is decomposed into a grid based representation, concepts such as objects or tracks do not exist. However, the object level representation may be demanded by the applications. To achieve this, a novel object detecting and tracking algorithm is presented. The prediction result of the tracking module is used as a form of feedback to the clustering module. The data association problem is taking into account by dealing with the ambiguous associations. Compared with the traditional joint probabilistic data association filter(JPDAF) method, the proposed algorithm demands less computational costs, so as to be suitable for environments with large amount of dynamic objects. The experiment result on the real data shows the effectiveness of the algorithm.