Implementación de Interfaz Gráfica para la Clasificación y Reconocimiento de Rostros Mediante el Procesamiento Digital de Señales

In this work, the analysis of two mainly facial recognition techniques, PCA and LDA, was developed in order to choose the most efficient technique for its future implementation in a graphical interface developed in the Python programming language. The facial recognition system will consist of two main stages, in the first stage the administrator can enter images captured from the videocamera to store the face of a new user; in the second stage the user can access the system by capturing a photograph of his face which will be compared with the databases previously created and the identification and photo of the user will be displayed. To perform a comparative analysis between the techniques mentioned above, a training stage was developed using ORL and MIT databases; observing the effect that produces the increase of the set of images in the final results. Subsequently, to compare the face of the input user with the faces present in the database, a classification stage was carried out using two KNN classification algorithms, one created by the autor and another developed by the Python SK-Learn library. Finally, the evaluation of the results was carried out using the SK-Learn library, this tool allows to obtain precisión percentages and mean Square error values of the developed models, these parameters are those that will allow an optimal selection of the model to be implemented. Keywords: PCA, LDA, Face recognition, SK-Learn, ORL, MIT, Detection, Classifier, Normalization.

• প্রধান লেখক: Nagua González, Bryan Manuel (author)
• বিন্যাস: bachelorThesis
• ভাষা: spa
• প্রকাশিত: 2019
• বিষয়গুলি: <INTERFAZ GRÁFICA><RECONOCIMIENTO DE ROSTROS><FOTOGRAFÍA><PROGRAMACIÓN PYTHON><TÉCNICA LDA>>
• অনলাইন ব্যবহার করুন: http://dspace.unl.edu.ec/jspui/handle/123456789/22600