Many Applications developed to recognition handwritten called Optical Character Recognition (OCR), who generally it only presented alphabet recognition. Javanese numbers (aksara wilangan) or Javanese characters are culture of Indonesia from great grandmother and must be knowed by rising genereation. The final project presented “Get Character Feature Handwriting Javanese Numbers Used Shadow Feature Extraction Method and Multi Layer Perceptron (MLP). The shadow feature method used to recognition characterstic of handwritten before it classification by MLP. Application test have two stage that is a sample training test 100 of data set and sample testing test 50 of data set. Percentage successed pattern concerning training samples 99% and error recognition 0,10%, whereas testing samples 90,8% and error recognition 9,2%.
Copyrights © 2015