Multi-Angle Facial Expression Recognition Method Based on Gabor Transform
Due to the instability of the face shape,there are significant differences in the visual images of faces from different observation angles.Under the influence of illumination change,facial expression posture and occlusion,it is difficult to accurately extract facial expression features,leading to low recognition accuracy.Therefore,a multi-angle facial expression recognition method based on Gabor transform was proposed.Through the eye location,the geo-metric preprocessing of multi-angle facial expression image was completed,so that the accuracy of facial expression recognition could be improved.Gabor transform was used to extract the feature of the multi-angle facial expression image.Moreover,the elastic template matching method was adopted to perform the elastic mesh matching on the fea-ture key points,and thus to calculate the cost function of the image.Finally,the K-nearest neighbor classification strategy was used to match and evaluate multi-angle facial expression images,thus completing the multi-angle facial expression recognition.Experimental results prove that the recognition time of the proposed method is less than 2s,and its recognition accuracy is close to 100%,so the method is more effective and accurate.
NETL
NSTL
万方数据
