摘要
手写体识别有着广阔的应用前景和很高的理论价值,其主要分为在线识别和离线识别两种。在分析局部线性嵌入算法LLE和文字识别过程的基础上,实现了整个手写文字识别过程,包括图像获取、预处理、特征提取、分类、LLE降维、识别输出等过程。通过文字的相容度试验,确定了本过程的有效性。
The recognition of the handwriting form has the broad application prospects and the very high theory val- ue. The recognition of the handwriting mainly divides into two kinds, namely on-line recognition and off-line recogni- tion character. Firstly, this paper studies the local linear embedding algorithm. Secondly, this paper designs to com- plete the entire process of character recognition. It includes image acquisition, pretreatment, feature extraction, clas- sification, LLE dimensionality reduction, the process of identifying, etc. Finally, the experimental data shows good re- suits.
出处
《河北联合大学学报（自然科学版）》
CAS
2012年第2期52-55,共4页
Journal of Hebei Polytechnic University:Social Science Edition
关键词
LLE算法
手写文字识别
降维
LLE algorithm
Handwriting recognition
Dimensionality reduction
