00000ctmaa22000003a 4500 UP-1685675941123923446 Buklod 20151030152225.0 aa rb |||1 u| ta 151030s xx d r |||| u| (iLib)UPTAC-00017783807 UPVTC eng LG 993.5 2007 C66 A46 Almo, Niel M. author. Character recognition "text to speech" Niel M. Almo [and] Mary Grace S. Comendador; John Paul T. Yusiong, adviser. 2007. [4], 51 leaves illustrations. Undergraduate thesis (B.S. Computer Science)--University of the Philippines, Tacloban. The computer world is constantly evolving each year. People make use of computers in many ways. Several applications known by man are implemented by introducing instructions to conventional computers. Until now, the target of computers to mimic the human brain and understand several tasks that humans don't know is still an on-going course of development. The main purpose of this research study is to use an approach using Artificial Neural Networks in recognizing alphanumeric characters. Character Recognition, in Computer Recognition, in Computer Science, is the process of examining printed characters on paper and determining which alphanumeric characters represent them by detecting patterns of dark and light. To make character recognition on computers possible, one has to use the concepts of Artificial Neural Networks. A training algorithm would be used in order for the network to learn characters, thus being able to recognize characters that were already introduced to them. Hence, the goal is to use an algorithm to train neural networks that is able to recognize characters and at the same time come up with a method to convert computer texts into speech. The BackPropagation algorithm has been used by the researchers to solve this problem. In the "speech" part of the system, the researchers used a microsoft agent to do the task. The researchers would therefore say that the BackPropagation algorithm is indeed a good algorithm when you are to deal with character recognition using Artificial Neural Network. It is because by using this algorithm, the researchers came up with a program that could recognize characters, provided that an intensive training was done on the network. Character recognition. Neural network. BackPropagation algorithm. Comendador, Mary Grace S. author. Yusiong, John Paul T. adviser. FI UP UPTAC UPTAC LG 993.5 2007 C66 A46 Thesis