Pengenalan Huruf Tulisan Tangan Berderau dan Terskala Berbasis Ekstraksi Ciri DCT dengan Menggunakan Jaringan Syaraf Probabilistik  

Oleh:

Sumarno, Linggo

Jenis: Article from Journal - ilmiah nasional - tidak terakreditasi DIKTI
Dalam koleksi: SIGMA: Jurnal Sains dan teknologi vol. 10 no. 2 (Jul. 2007), page 185-197.
Topik: Handwritten Letter; Noisy; Scaled; DCT; Probabilistic Neural Network

Ketersediaan

Perpustakaan Pusat (Semanggi) Nomor Panggil: SS25.6 Non-tandon: 1 (dapat dipinjam: 0) Tandon: tidak ada

Lihat Detail Induk

Isi artikel

This paper proposes a system of off-line recognition of noisy and scaled handwritten letters based on DCT (Discrete Cosine Transform), which is used extensively in lossy compression of data. The system employs Probabilistic Neural Netwok, which is used in many engineering problems and pattern recognition. Simulation results on existing database, which was created by a single writer, indicate that the system is adequate to be used up to certain levels of noise and scaling. At noise level 0% and scale 1, the system is able to get recognition rate up to 91%, whereas at noise level 0% and scale 0.7, as well as at noise level 10% and scale 0.8, the recognition rates are 84.6% and 79.9%, respectively.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali