Entropy-based feature extraction algorithm for stone carving character detection

The Journal of Engineering

Changan Yuan¹  Xiao Qin²  Xutao Chu³ 

[1] Massey University , Auckland , New Zealand;School of Computer and Information Engineering , Guangxi Teachers Education University , Nanning , People's Republic of China

关键词: historical values;    text detection;    print fonts;    stone carving character detection;    semantic recognition;    stone texts;    natural scenes;    stone text extraction;    entropy-based feature extraction algorithm;    machine learning algorithms;    printed text detection;    h;    -written fonts;    cultural values;    h;    written detection;   

DOI  :  10.1049/joe.2018.8318

学科分类：工程和技术（综合）

来源: IET

【 摘 要 】

Text detection is the premise of semantic recognition of natural scenes. At present, many studies only focus on print fonts or hand-written fonts. There are many stone texts in China that reflect historical and cultural values but are difficult to identify. Detecting these stone carving texts in natural scenes is more difficult than detecting them in handwritten or printed texts. In this study, the entropy-based feature extraction algorithm is proposed to extract stone texts in natural scenes. Compared with other excellent machine learning algorithms, it has achieved better results.

【 授权许可】

CC BY