五邑大学学报(自然科学版)
五邑大學學報(自然科學版)
오읍대학학보(자연과학판)
JOURNAL OF WUYI UNIVERSITY(NATURAL SCIENCE EDITION)
2015年
2期
49-54
,共6页
表情识别%纹理特征%改进局部三值模式算法%稀疏表达分类器
錶情識彆%紋理特徵%改進跼部三值模式算法%稀疏錶達分類器
표정식별%문리특정%개진국부삼치모식산법%희소표체분류기
facial expression recognition%texture features%ILTP%SRC
为了提高实际应用中的人脸表情识别率,本文提出了改进局部三值模式算法(ILTP),并结合稀疏表达分类器(SRC)组成新的算法应用于人脸表情识别.该算法首先利用 ILTP 算法对人脸表情图像进行特征提取,然后将得到的图像顶层特征数据和图像底层特征数据作为SRC 的输入,从而完成人脸表情分类.基于 JAFFE 数据的实验结果表明:改进算法的人脸表情识别率达70.48%,具有较高的可行性.
為瞭提高實際應用中的人臉錶情識彆率,本文提齣瞭改進跼部三值模式算法(ILTP),併結閤稀疏錶達分類器(SRC)組成新的算法應用于人臉錶情識彆.該算法首先利用 ILTP 算法對人臉錶情圖像進行特徵提取,然後將得到的圖像頂層特徵數據和圖像底層特徵數據作為SRC 的輸入,從而完成人臉錶情分類.基于 JAFFE 數據的實驗結果錶明:改進算法的人臉錶情識彆率達70.48%,具有較高的可行性.
위료제고실제응용중적인검표정식별솔,본문제출료개진국부삼치모식산법(ILTP),병결합희소표체분류기(SRC)조성신적산법응용우인검표정식별.해산법수선이용 ILTP 산법대인검표정도상진행특정제취,연후장득도적도상정층특정수거화도상저층특정수거작위SRC 적수입,종이완성인검표정분류.기우 JAFFE 수거적실험결과표명:개진산법적인검표정식별솔체70.48%,구유교고적가행성.
In order to improve the facial expression recognition rate in practical application, an improved local ternary patterns (ILTP) algorithm was proposed on the basis of the local ternary patterns (LTP) algorithm,and was combined with sparse representation-based classifier (SRC) to form a new algorithm to be applied to human facial expression recognition. Then facial expression features are extracted by ILTP algorithm, and the features are treated as the input of the SRC to complete facial expressions recognition. Experimental results based on JAFFE database prove that the new algorithm can get a facial expression recognition rate of 70.48% and is highly feasible.