Off-line Arabic handwriting recognition system based on ML-LPQ and classifiers combination
Aicha Korichi, Oussama Aiadi, Belal Khaldi, Sihem Slatnia, Mohammed Lamine Kherfi · 2018
Recently, Handwriting recognition has received a growing interest from researchers. One of the challenging research areas in this field is Arabic handwriting recognition because of the intrinsic characteristics of Arabic language. Despite the growing researches in the last years, many efforts remain to be done. In this paper, we propose an Arabic handwriting recognition system based on a Multi-scale Local Phase Quantization (ML-LPQ) descriptor. In order to improve the recognition rate, we have combined three classifiers at the decision-level namely Support Vector Machine, Naïve Bayes and K-Nearest Neighbor. To conduct experiments, we have introduced a new database that assembles concepts from the computer science field. Experimental results, given at the end of the paper, have demonstrated the efficiency of the proposed method and promising results have been achieved.