欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
Zernike不变矩具有对噪声不敏感,正交等特性,是表情的有效表征方法,高阶Zernike矩包含更多图像信息,对表情分类的作用更大。但是高阶矩的计算复杂度很大,很难达到快速表情识别的要求。本文利用小波变换对表情图像进行多尺度分析,从低频子图像中计算其Zernike矩作为判别特征进行表情识别。通过小波变换,一方面可以对图像降维,降低计算复杂度;另一方面,小波变换的去噪性能使得识别效果更好。实验表明,基于多尺度分析Zernike矩特征的方法优于单独使用小波变换或Zernike矩特征方法的识别效果。
英文摘要：
Zernike invariant moments are orthogonal and insensitive to noise,and they are efficient for facial expression represen- tation, especially the moments with high order. Although high order moments are more useful to classification, the computation is com- plex. Therefore, we present a novel feature extraction method based on wavelet transform and Zernike moments. The facial images are de- composed using wavelet transform and Zernike moments of the low frequency domain are treated as feature vectors for facial expression recognition. Wavelet transform is used to reduce the dimensionality of the facial images and weaken the noise, so that the moments can be computed easily, and higher recognition accuracies are achieved. Experiment results indicate that this method outperforms the methods which use wavelet transform or Zernike moments alone.
同期刊论文项目
同项目期刊论文
A Novel Coupled Bi-directional Diffused Image Inpainting Model with Morphological Invariant Property
期刊信息
位置：