PENGENALAN KARAKTER TULISAN TANGAN MENGGUNAKAN AFFINE MOMENT INVARIANT DAN SELF ORGANIZING MAPS
Daftar Isi:
- Handwriting characters recognition is an area of recognition process that has become an interesting research subject. The problem in handwriting character recognition is the characteristic of each character that is similar so that it is often incorrectly recognized by the computer. To overcome these problems developed handwriting character recognition software that can recognize characters well. In this research, software that can perform handwriting recognition developed using Affine Moment Invariant (AMI) method and Self Organizing Maps (SOM). The input of this software is a handwritten character image. The image will go through a pre-processing process using binarization. The features of the binary image be captured using AMI which yield 4 values and trained using SOM to get the best weights used during the character recognition. The result of handwriting character recognition using SOM is of the 1240 character images tested obtained an accuracy of 86.75%.