PERANCANGAN PERANGKAT LUNAK UNTUK PENGENALAN KARAKTER ASCII DARI GAMBAR BITMAP MENGGUNAKAN JARINGAN SARAF TIRUAN METODE PROPAGASI BALIK

Abstract

ANN is a method in soft computing that imitate the structure of biological nervous, where multiple nodes communicate with each other through synapses that interconnect them. That method can be utilized for pattern recognition processes, such as ASCI Icharacter recognition.This paper is aimed to develop ASCII character recognition using ANN method. Multilayer neural network is adopted in ANN with back propagation algorithm. A mapping method is used for preprocessing to segment a character image to be processed by ANN.The experimental results show that among all of successful segmented characters of all the training data having the different size of font, the system successfully recognizes characters up to 81.74 %. For the training data with different sizes and fonts data, the system recognizes the characters with an accuracy close to 55.74%. Among the unsuccessful segmented characters of all training data having different sizes of font, the system recognizes characters with an accuracy 67.53%. For the different sizes and fonts data, the system recognizes the characters with an accuracy close to 46.65%.