Facial recognition system applied to multipurpose assistance robot for social human-robot interaction (MASHI)
Face recognition is one of the key areas in the field of pattern recognition and artificial intelligence (AI). It has been used in a wide range of applications, such as identity authentication, biometrics, and surveillance. Image data is high dimensional in the face recognition area, so requires a c...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2017 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/108555 |
| Acceso en línea: | https://hdl.handle.net/2117/108555 |
| Access Level: | acceso abierto |
| Palabra clave: | Robotics Robòtica Àrees temàtiques de la UPC::Informàtica |
| Sumario: | Face recognition is one of the key areas in the field of pattern recognition and artificial intelligence (AI). It has been used in a wide range of applications, such as identity authentication, biometrics, and surveillance. Image data is high dimensional in the face recognition area, so requires a considerable amount of computing resources and time for recognition. Research effort has been developed in this way, and nowadays many algorithms are available for solving this problem in Computer Vision. The main goal of this project is to improve the capabilities of the MASHI robot, endowing it for more interaction with humans, and add new functionalities with the components that the robot has. FISHERFACES, a popular technique for facial recognition is the one chosen to be implemented in our application. This work studies the mathematical fundamentals of this technique to understand how information is processed to perform face recognition. Then, some tests have been performed to check the reliability of the application with several databases of facial images. In this way, it is possible to determine the strengths and weaknesses of the algorithm to be implemented in our robot. This work introduces an implementation based on Python using the OpenCV library. The characterization of hardware and the description of software is presented. Next, results, limitations, future works, and conclusions over the job development are presented. |
|---|