Diseño de una técnica para el reconocimiento de imágenes de signos basados en el alfabeto de la lengua de signos peruana (LSP) utilizando el descriptorblurred shape model

• Autor: Herrera Querevalú, Edson Martin Jair
• Formato: tesis de grado
• Fecha de Publicación: 2014
• Institución: Universidad Nacional de Trujillo
• Repositorio: UNITRU-Tesis
• Lenguaje: español
• OAI Identifier: oai:dspace.unitru.edu.pe:20.500.14414/11305
• Enlace del recurso: https://hdl.handle.net/20.500.14414/11305
• Nivel de acceso: acceso abierto
• Materia: Reconocimiento de imágenes; LSP; Descriptorblurred shape model

The present work had been special related with the fact of those people with hearing problems that need visual signs (in this case, signs that belong to the “Lengua de Signos Peruana” LSP) to communicate with other people; that is why from this fact I presented a study about the design of a technic for the recognition of signs based in the “Lengua de Signos Peruana” (LSP) using the Blurred Shape Model (BSM) descriptor. For the design of this technic, we used image processing algorithms in every phase of the technic, which were: pre-process, segmentation, description and recognition, which together generated the technic to design. To determine what algorithms we had to use in each phase, first of all, they had to analyze according to varied evaluation criteria, which they went according the needs of this work and determined their selection and utilization, once they had determined the algorithm to use, the next step was the technic’s design, in which specifies and details the process performed for each phase, using the selected algorithms. To measure the obtained results, we used the Confusion Matrix, which allowed to calculate the individual effectiveness of recognition for each kind of sign, and the average effectiveness of the technique. Finally we concluded that the proposed technique is a powerful tool for recognizing images of signs based “Lengua de Signos Peruana”