Desarrollo de un dispositivo de identificación de números y letras empleando hardware y software libre orientado a la enseñanza de niños de Educación Básica.

Developing a device to identify numbers and letters, by free Software and Hardware, aimed at teaching children into elementary Education, and is aimed at vulnerable populations from the spaces of Social Projection was the aim of this research, since according to the Constitution of the Republic of Ecuador, education is a right of people throughout their lives, which led to an investigation based on this need. The device processes artificial vision, sending it in real time to an interface, a place where the image is classified to recognize numbers or letters as identified. For this, neural networks were used, the same ones that were trained using free software, for both uppercase and lowercase letters and numbers, thus allowing the system not to be saturated with a single network. Once the weights given by the training of the neural networks have been obtained, the algorithm is implemented for its subsequent validation. From the tests carried out it can be determined that the module has an accuracy and reliability of 95 to 98% for its use, in addition it was experimented with elementary Education teachers adjusting to their needs and requirements. The device is easy to use for teachers, or people who have basic computer knowledge, since its interface is very similar to that of a desktop computer, it also has the advantage of being portable and not depending on additional resources such as the internet.

• Hovedforfatter: Hugo Livardo, Alomía Landázuri (author)
• Andre forfattere: David John, Cedeño Arcos (author)
• Format: bachelorThesis
• Sprog: spa
• Udgivet: 2020
• Fag: VISIÓN ARTIFICIAL; HARDWARE LIBRE; SOFTWARE LIBRE; REDES NEURONALES; ELECTROMECÁNICA
• Online adgang: http://repositorio.utc.edu.ec/handle/27000/8034