基于改进的Gabor和ADABOOST的人脸表情识别

Facial Expression Recognition Based on Improved Gabor Transformation and Adaboost Algorithm

介绍了受到不同程度形变和光照影响的人脸,如何提高表情图像识别效率和识别准确率的方法。采用Log_Gabor小波变换方法对输入图像进行滤波处理,并采用AdaBoost方法对人脸表情进行识别处理的方法。实验表明人脸表情在形变和光照影响的情况下,对人脸表情识别准确率依然达到良好的效果,且提高了计算速度。

The recognition efficiency and accuracy of facial expressions influenced by transformation and sunlight to different degrees were described and introduced. Log-Gabor small-wave transformation methods were used to filter input images. Adaboost was also adopted to recognize facial expressions. The research indicated that accuracy rate of recognizing facial expressions was achieved when the facial expressions were transformed or influenced by the sunlight. Calculating speed was also improved.