Reconocimiento facial de personas basado en el análisis de componentes principales PCA

• Autores: Collantes Arana, Alan Alfredo, Sari Bustos, Dany Richard
• Formato: tesis de grado
• Fecha de Publicación: 2013
• Institución: Universidad Nacional de Trujillo
• Repositorio: UNITRU-Tesis
• Lenguaje: español
• OAI Identifier: oai:dspace.unitru.edu.pe:20.500.14414/8553
• Enlace del recurso: https://hdl.handle.net/20.500.14414/8553
• Nivel de acceso: acceso abierto
• Materia: Reconocimiento facial, Análisis PCA

In the present thesis, there carried out the study of the mathematical technique of the Analysis of Principal Components – PCA. It was used for the persons' facial recognition and gotten to decrease the rate of mistake. To get this aim, it used images that were taken of the base faces of the Olivetti Research Laboratory, which have low resolution and are in grayscale._x000D_ With the purpose of achieving better results in the recognition, the first step that was realized is to apply a series of Gabor's Filters, in different frequencies and orientations, searching with this to highlight the most relevant characteristics of the face in the images._x000D_ On these images obtained of Gabor's Filter, the second step was realized with the application of the technique of the Analysis of Principal Components - PCA, the purpose of decreasing the dimension of the set of patterns that they represent the faces of training, where there are obtained the " Vectors of Characteristics " of the faces, also known as "Eigenvectores". Finally, the third step, our image of proof is transformed to the new dimension, through the Eigenvectores, and in this way is obtained the recognition of the subject._x000D_ In addition, it realized the comparison of the results that were obtained of the combination of Gabor's Filters and the Analysis of Principal Components- PCA, with the results of the application of the method of the Analysis of Principal Components - PCA, demonstrating in such a way that the achievement of better results as to the rate of mistake of the recognition