Public databases

Main publications

• B. Cancela, A. Iglesias, M. Ortega, M. G. Penedo, "Unsupervised Trajectory Modelling using Temporal Information via Minimal Paths", IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2014), Columbus, Ohio, June 2014.
  [Abstract] [PDF]
  This paper presents a novel methodology for modelling pedestrian trajectories over a scene, based in the hypothesis that, when people try to reach a destination, they use the path that takes less time, taking into account environmental information like the type of terrain or what other people did before. Thus, a minimal path approach can be used to model human trajectory behaviour. We develop a modified Fast Marching Method that allows us to include both velocity and orientation in the Front Propagation Approach, without increasing its computational complexity. Combining all the information, we create a time surface that shows the time a target need to reach any given position in the scene. We also create different metrics in order to compare the time surface} against the real behaviour. Experimental results over a public dataset prove the initial hypothesis' correctness.
• B. Cancela, M. Ortega, M. G. Penedo, "Multiple Human Tracking System for Unpredictable Trajectories", Machine Vision and Applications, 25(2), 511-527, 2014.
  [Abstract] [PDF]
  Tracking multiple objects into a scene is one of the most active research topics in computer vision. The art of identifying each target within the scene along a video sequence has multiple issues to be solved, being collision and occlusion events the most challenging ones. Because of this, when dealing with human beings, it is often very difficult to obtain a full body image, which introduces complexity in the detection. Thus, head-shoulder omega shape becomes a powerful tool to perform the human detection. Most of the contributions to this field involve a detection technique followed by a tracking system based in the omega shape features. No collision and occlusion events are provided, making the systems have poor results in challenge environments. Based in these works, we present a novel methodology for providing a full tracking system. Different techniques are combined to both detect, track and recover target identifications under uncontrolled scenarios. Experimental results into challenging sport scenes show the performance and accuracy of this technique. Also, the system speed opens the door for obtaining a real-time system using GPU programing in standard desktop machines, being able to be used in higher-level human behavioral systems, with multiple applications.
• B. Cancela, M. Ortega, M. G. Penedo, J. Novo, N. Barreira, "On the Use of a Minimal Path Approach for Target Trajectory Analysis", Pattern Recognition, 46 (7), 2015 - 2027, 2013.
  [Abstract] [PDF]
  Vision-based action recognition has multiple applications, mainly focused in video surveillance systems. The art of labeling each target behavior in crowded scenarios is a complicated field since usually we do not have visual confirmation of the parts of a target to infer its behavior. Thus, trajectory analysis becomes a good choice to try to infer knowledge about target movements. Most of the contributions to this field involve a training period in which we obtain information a priori about the environment, storing a dataset with all the possible usual routes. Based in the minimal path theory using geodesic active contours, we present a novel architecture where no initial information about the scene is needed, while it is possible to include it if necessary to specify constraints. A multiple-target tracking methodology is used in order to achieve all target routes, detecting both occlusions and target collisions. Experimental results in two different application domains show the performance and flexibility of this method, being able to be used in multiple trajectory analysis problems.
• B. Cancela, M. Ortega, A. Fernández, M. G. Penedo, "Hierarchical framework for robust and fast multiple-target tracking in surveillance scenarios", Expert Systems with Applications, 40, 1116 - 1131, 2013.
  [Abstract] [PDF]
  Multiple-target tracking is a challenging field specially when dealing with uncontrolled scenarios. Two common approaches are often used, one based in low-level techniques to detect each object size, position and velocity, and other based in high-level techniques that deal with object appearance. None of these methods can deal with all possible problems in multiple-target tracking: environment occlusions, both total and partial, and collisions, such as grouping and splitting events. So one solution is to merge these techniques to improve their performance. Based in an existing hierarchical architecture, we present a novel technique that can deal with all the mentioned problems in multiple tracking targets. Blob detection, low-level tracking using adaptive filters, high-level tracking based in a fixed pool of histograms and an event management that can detect every collision event and performs occlusion recovery are used to be able to track every object during the time they appear within the scene. Experimental results show the performance of this technique under multiple situations, being able to track every object in the scene without losing their initial identification. The speed processing is higher than 50 frames, which allows it to be used under real-time scenarios.