Visual tracking of an object

• US 9,886,768 B2
• Filed: 04/25/2014
• Issued: 02/06/2018
• Est. Priority Date: 04/26/2013
• Status: Active Grant

• Current Assignee: Centre National De La Recherche Scientifique, Sorbonne Universite
• Original Assignee: University Pierre et Marie Curie
• Inventors: Regnier, Stphane, Benosman, Ryad, Pacoret, Ccile, Siohoi, Ieng, Zhenjiang, Ni
• Primary Examiner(s): Bali, Vikkram
• Application Number:
  US14/786,577
• Publication Number:
  US 20160086344A1
• Time in Patent Office: 1,383 Days
• Field of Search: None
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06T 2200/28 involving image processing hardware
• CPC Class Codes: G06T 2207/30108 Industrial image inspection
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/254 involving subtraction of images
• CPC Class Codes: G06T 7/277 involving stochastic approaches, e.g. using Kalman filters
• CPC Class Codes: G06T 7/77 using statistical methods

Abstract

Method for visual tracking of at least one object represented by a cluster of points with which information is associated, characterized in that it includes steps to: receive (E1) data representing a set of space-time events, determine (E2) the probability that an event in the set belongs to the cluster of points representing the at least one object, for each event in the received set, determine (E3) whether or not an event belongs to the cluster of points as a function of the determined probability for the event considered, for each event in the received set, update (E4) information associated with the cluster of points for at least one object, for each event for which it was determined in the previous step that it belongs to the cluster of points, calculate (E4, E5) the position, size and orientation of the at least one object as a function of the updated information.

Claims

1. Method for visual tracking of at least one object represented by a cluster of points with which information is associated, wherein the method includes steps to:

receive (E1) data representing a set of space-time events from an asynchronous sensor

(90);

determine (E2) the probability that an event in the set belongs to the cluster of points representing the at least one object, for each event in the received set;

determine (E3) whether or not an event belongs to the cluster of points as a function of the determined probability for the event considered, for each event in the received set;

update (E4) information associated with the cluster of points for at least one object, for each event for which it was determined in the previous step that it belongs to the cluster of points, a position of the at least one object being updated as a function of the position at a previous time of the at least one object and the event determined at the previous step as belonging to the cluster of points representing the at least one object;

calculate (E4, E5) the position, size and orientation of the at least one object as a function of the updated information.

2. Method for visual tracking of at least one object according to claim 1, wherein the determination (E3) of whether or not an event belongs to the cluster of points includes a comparison of the determined probability for the considered event with a predetermined threshold.

3. Method for visual tracking of at least one object according to claim 1, wherein size and orientation of the at least one object are calculated (E5) by modelling the object by an ellipse and calculating its half axes and its orientation.

4. Method for visual tracking of at least one object according to claim 1, wherein the method also comprises determination of the speed of the at least one object.

5. Method for visual tracking of at least one object according to claim 1, wherein the reception, determination, update and calculation steps are carried out iteratively.

6. Method for visual tracking of at least one object according to claim 1, wherein the method also comprises the step to display (E6) data representing the position, size and orientation of the at least one object calculated as a function of the updated information.

7. A non-transitory computer readable storage medium comprising instructions for a computer program that, when executed, performs the steps in the method according to claim 1.

8. Device for visual tracking of at least one object represented by a cluster of points associated with information, wherein the device comprises:

means of receiving data representing a set of space-time events;

means of determining the probability that an event in the set belongs to the cluster of points representing the at least one object, for each event in the set received by the reception means;

means of determining whether or not an event belongs to the cluster of points as a function of the probability determined for the event considered, for each event in the set received by the reception means;

means of updating information associated with the cluster of points for the at least one object, for each event for which it has been determined that it belongs to the cluster of points, a position of the at least one object being updated as a function of the position at a previous time of the at least one object and the event determined at the previous step as belonging to the cluster of points representing the at least one object;

means of calculating the position, size and orientation of the at least one object depending on the updated information.