Generating an Arabic Calligraphy Text Blocks for Global Texture Analysis

This paper objective is to improve the current method for generating an Arabic Calligraphy text blocks. We test on seven types of Arabic Calligraphy text. We apply projection profiles and a proposed filter to discriminate each line of the Arabic Calligraphy scripts. After performing text detection, skew correction, text and line normalization subsequently, we generate Arabic Calligraphy text blocks for global texture analysis purposes. We compare our proposed filter with current method and median filter. The results show that the proposed filter is outperformed. The proposed method can be further improved to boost the overall performance.