APLIKASI PENGENALAN KARAKTER TULISAN TANGAN DENGAN EKSTRAKSI CIRI DWT

Yusten Wuntoro, Chairisni Lubis, Agus Budi Dharmawan

Abstract


Some research indicates that every person's handwriting is unique and can be recognized by special software. To translate handwritten characters then made an application to the method of DWT (Discrete Wavelet Transform). The process is done in recognition of handwritten characters is done transferring pictures to a computer handwritten, pre-processing by using grayscale, thresholding, normalization size, thinning, edge detection Sobel, slant angle that is used when a character italics, the feature extraction Discrete Wavelet Transform is used and recognition using Euclidean Distance and Manhattan Distance. Test results from all capital letters to 260 data with 208 training data and 52 55% of the successful introduction of data on Haar lv 1, 52% in Haar lv 2, 48% on Daubechies lv 1, and 67% on Daubechies lv 2.


Refbacks

  • There are currently no refbacks.