KLASIFIKASI NAMA OBAT TULISAN TANGAN DOKTER DENGAN METODE GLCM DAN BACKPROPAGATION NEURAL NETWORK

Arrahman Arrahman
Purwanto Purwanto
Pulung Nurtantio

Jurnal Teknologi Informasi Cyberku
Vol 12 No 2 (2016): Jurnal Teknologi Informasi CyberKU Vol. 12, no 2

Pharmaceutical personnel in the work in general is always associated with the reading prescription, where the required accuracy, speed, and accuracy in reading prescription to avoid medication errors. This research show how to classify the doctor's handwriting drug name. Research conducted by the image processing prescription taken by scanner. Then the image manually cropped to take 200 drug names. Refining the drug name image has done twice with median filter and wiener filter, then dilation and erosion , feature extraction with GLCM (Grey-Level Co-occurance matrix) methods to obtain data sets that will be classified by the software RapidMiner. From the test we find that Backpropagation Neural Network had more accurate than Naive Bayes and C 4.5.