期刊名称:International Journal of Computer Science and Network Security
印刷版ISSN:1738-7906
出版年度:2017
卷号:17
期号:4
页码:352-360
出版社:International Journal of Computer Science and Network Security
摘要:In this study, the Local Gabor Binary Pattern (LGBP) algorithm were used for facial expression recognition of emotion (happiness, sadness, anger, disgust, surprise and fear). As the topic of finding a strong feature has been studied by many research, in this work, we decided to focus on improving performance and features extracted more important features to get more accuracy of recognition.
The investigation came to the conclusion that considering the different weight matrix in the detection of facial expressions can be important parts of the face become more prominent. Therefore, the input image is partitioned into 9 equal area and extract the LGBP features. Then the entropy in each of those areas multiplied with the matrix derived features in areas where in areas where entropy is higher, like around the eyes, eyebrows and mouth vector Features with more effective and more affect Classifier.
KNN classifier is used in this research, which is used as a weighting fuzzy and without using fuzzy weighting. We reached in fuzzy mode 1.66 percent more accurate than other. Test results also confirms influence the partitioning of the area and using entropy weighting in areas.