Tracking of multiple objects under partial occlusion

Abstract

The goal of multiple object tracking is to find the trajectory of the target objects through a number of frames from an image sequence. Generally, multi-object tracking is a challenging problem due to illumination variation, object occlusion, abrupt object motion and camera motion. In this paper, we propose a multi-object tracking scheme based on a new weighted Kanade-Lucas-Tomasi (KLT) tracker. The original KLT tracking algorithm tracks global feature points instead of a target object, and the features can hardly be tracked through a long sequence because some features may easily get lost after multiple frames. Our tracking method consists of three steps: the first step is to detect moving objects; the second step is to track the features within the moving object mask, where we use a consistency weighted function; and the last step is to identify the trajectory of the object. With an appropriately chosen weighting function, we are able to identify the trajectories of moving objects with high accuracy. In addition, our scheme is able to handle partial object occlusion.

Tracking partially occluded objects with centripetal active contour

“…A boundary template generated by S-ACES in the first frame is used for the matching. An advantage of the present method compared to the Kalman [11] and those present in [8,24] is the higher speed. Another advantage is that the geometric information extracted from the objects in a frame is useful to determine the semantic of the scene.…”

“…Another advantage is that the geometric information extracted from the objects in a frame is useful to determine the semantic of the scene. Also, the present method is capable of distinguishing objects with visually similar shapes while the methods from [8,24] will miss the target in such a situation. However, a disadvantage of the present approach is that it cannot handle right now a tracking of multiple targets as [8,24] can do.…”

“…In [8], Han et al have proposed method for tracking multiple objects under partial occlusion. In their work the object is represented with multiple features.…”

“…A number of tracking algorithms are available in the literature [1,6,8,11,13,14,20,24]. Generally, tracking systems face two basic obstacles: motion detection, and matching.…”

Adaptive Kalman Filter Applied to Vision Based Head Gesture Tracking for Playing Video Games

“…To deal with problems raised due to occlusion, Han et al [20] applied feature tracking in three steps: initially detecting the moving object, then tracing feature points within the object mask and finally estimating the trajectory of the object. Hua et al [21] proposed a way to combine occlusion and motion estimation with a tracking-by-detection approach.…”