Implementación de una aplicación de tracking, con el fin de conocer la posición espacial de una persona, mediante el uso de diferentes cámaras y técnicas de visión artificial

Abstract

[ES] The following project aims to implement and analyze different monitoring techniques for one or more users in a known space, thus returning 3D data about their location and movement. In turn, a communication method will be developed through the network, between applications and devices to make use of said movement data.

In order to compare different methodologies, three tracking applications will be implemented using different hardware and artificial vision techniques. On the one hand, a semi-professional equipment will be used that will consist of a 3D camera, which has different sensors that allow it to build a depth mapping and in this way obtain the coordinates of the users of the room in a precise way.

On the other hand, a much more everyday model will be used that will use ordinary webcams to track the user. In this way, two applications will be implemented using webcams:

An application with two cameras, which, through facial recognition techniques, correct calibration and triangulation, is able to obtain results on the user's location in a similar way as with the 3D camera.

Another application that uses a single camera, and in a similar way to the previous one, through facial recognition techniques, tracks the user, but this time to calculate the coordinates of the user, different methodology will be used, which will be based on the knowledge of certain parameters initials. To obtain these parameters, an algorithm will be implemented that, based on the interaction with the user and in addition to certain notions about anatomical proportions, are calculated precisely.

The objective of the project is not only the implementation of the different tracking applications, but also the results obtained and the advantages and disadvantages of each model will be compared. For this, parameters such as the precision of the model, the processing time between frames or the price will be analyzed.

[EN] The following project aims to implement and analyze different monitoring techniques for one or more users in a known space, thus returning 3D data about their location and movement. In turn, a communication method will be developed through the network, between applications and devices to make use of said movement data.

In order to compare different methodologies, three tracking applications will be implemented using different hardware and artificial vision techniques. On the one hand, a semi-professional equipment will be used that will consist of a 3D camera, which has different sensors that allow it to build a depth mapping and in this way obtain the coordinates of the users of the room in a precise way.

On the other hand, a much more everyday model will be used that will use ordinary webcams to track the user. In this way, two applications will be implemented using webcams:

An application with two cameras, which, through facial recognition techniques, correct calibration and triangulation, is able to obtain results on the user's location in a similar way as with the 3D camera.

Another application that uses a single camera, and in a similar way to the previous one, through facial recognition techniques, tracks the user, but this time to calculate the coordinates of the user, different methodology will be used, which will be based on the knowledge of certain parameters initials. To obtain these parameters, an algorithm will be implemented that, based on the interaction with the user and in addition to certain notions about anatomical proportions, are calculated precisely.

The objective of the project is not only the implementation of the different tracking applications, but also the results obtained and the advantages and disadvantages of each model will be compared. For this, parameters such as the precision of the model, the processing time between frames or the price will be analyzed.