object tracking

The term 'object' is general and also includes persons or animals, depending on the application scenario. The task is particularly challenging when there are many objects of interest within the same scene. So, for example, objects are vehicles in traffic or boats at the entrance to a harbor, tracked animals include bees, and the most frequently tracked people are pedestrians or customers moving in shops.

A special case is person tracking when exploiting characteristics peculiar of this subject, like face, articulated part, body shape, gait.


The design of a multi-object detection and tracking system involves two main phases: object detection and object association. In the first phase, the desired objects are detected in each frame of the video stream. The quality of the detection directly influences the tracking performance. The second stage involves the association of the detected objects in the current frame with previous ones to obtain their trajectories. High accuracy in object detection results in fewer missed detection and ultimately produces less fragmented tracks.


  1. J.C. SanMiguel, J. Munoz, F. Poiesi. Detection-aware multi-object tracking evaluation, IEEE International Conference on Advanced Video and Signal Based Surveillance - AVSS 2022

  2. S. Messelodi, C.M. Modena, V. Ropele, and S. Marcon ans M. Sgrò. A Low-Cost Computer Vision System for Real-Time Tennis Analysis. 20th International Conference on Image Analysis and Processing - ICIAP 2019 [doi]

  3. Messelodi, C.M. Modena. Scene Text Recognition and Tracking to Identify Athletes in Sport Videos. Multimedia Tools and Applications 63(2):521-545, 2013.


  1. S. Messelodi, C.M. Modena, M. Zanin. A computer vision system for the detection and classification of vehicles at urban road intersections. Pattern Analysis and Applications 8(1-2):17-31, 2005 [doi]

  2. M. Rossi, A. Bozzoli. Tracking and Counting Moving People, International Conference on Image Processing - ICIP 1994 [doi]