Abstract:Ear biometric has the basic characteristics,such as universality,distinctiveness,stability and collectibility.The non-disturbed recognition can be easily realized by ear biometric or fusion of ear and face biometrics.As a novel biometric identification,ear recognition is a very active research project in the field of pattern recognition.However,ear recognition result is usually affected by the head posture of image,illumination variations,occlusion and other factors,which makes the process of recognition more complicated and challenging.In this paper,a novel ear recognition approach based on histogram of oriented gradient feature extraction is studied.The ear recognition scheme by combining histogram of oriented gradient features with sub-region fuzzy fusion is presented.The ear image is divided into a number of sub-regions,histograms of oriented gradient features of different sub-regions are extracted separately,and the fuzzy membership matching fusion strategy is introduced to obtain the last classification label.Furthmore,the issue on ear recognition under local occlusion based on histogram of oriented gradient is discussed in detail.The recognition contribution rate for each sub-region is defined and analyzed,and the accumulated occlusion is investigated,which can give a reference for determining the effective identify regional block.The maximum occluded range and effective identify regional block can effentively identify the human ear images.Therefore,the standard of the ear images whether retained or discarded can be determined.Experimental results show that ear recognition approach based on histogram of oriented gradient features is practicable and effective.

Ear recognition with histogram of oriented gradient features