Visual Multiple Object Tracking (VMOT) is an important computer vision task which has gained increasing attention due to its academic and commercial potential. There are many different approaches have been proposed to solve the problem. Compared with single object tracking which focuses on appearance model, motion model and other factors, multiple object tracking shares these common challenges, and has some more challenging tasks to be tackled, such as identifying among multiple objects, frequent occlusion due to the crowd, initializing and terminating of the object tracking, et al, which making it difficult to get a comprehensive understanding of this problem. In this paper, we try to make systematic review on VMOT. We review the recent advances in various aspects and introduce the classification of each aspect in detail. The main aspects include observation model and tracking algorithm. In each aspect, the existing methods are classified into different groups, and each group is discussed in details for their principles, advantages and drawbacks. This review work can provide researchers, especially new comers to the topic of VMOT, a general understanding of the state-of-the-arts, and help them to comprehend the aspects of a VMOT system and the inter-connection of these aspects.