On Image Normalization in Ear Recognition
Zhichun Mu · Control Engineering of China · 2008
To the problem that image needs to be normalized automatically and the normalization results be valid and reasonable for feature extraction in ear recognition,a method is proposed to normalize the images to a certain size and direction according to the long axis of the earlap.Active shape model is used to search the earlap automatically and the model is improved to adjust the initial position.Experiment results show that the improvement can make the model converge to the object much faster and more exactly,moreover the normalization results are reasonable and good for later feature extraction.