Name Essays (3218 words) - Auditory System, Perception, Surveillance

Name Essays (3218 words) - Auditory System, Perception, Surveillance Name Professor Course Date Abstract The recognition of the ear is determined by the ear biometrics with respect to the physiological as well as the behavioral characteristics of the ear. The stable biometric nature of the ear draws the interest in many scholars in conducting a number of researches on the recognition capability plus functionality of the ear. Other biometric aspects of the other body parts indicate that the biometrics of the face and the figures are less accurate in comparison with the biometrics of the ear. The ear biometric is a good component of recognition that other body parts because the ear biometric is fixed. Even though in industrial and academic fields these have been the major recognition concepts, the ear recognition stands out to be the feature which gives the accurate and distinct recognition when compared to the rest of the body parts. This research paper outlines the use of deep counter as an alternative technique for edge detection to the capability of recognition of the ear. The main stages of deep counter methodology of studying the ear recognition ability are the ability of the ear to conduct prior processing of information, skin detection, contrast enhancements and size normalization. The CHAINLETS help to extract all the ear features then helping the ear to match all the information for proper identification. This method of ear detection is most applicable in an unconstrained channel of ear recognition. The results of the experiment indicate that the approach of Deep Counter and CHAINLETS for Ear recognition show the best result comparing with another state of the art base descriptors. Introduction According to Moreno B., Sanchez A., Velez J. (1999) ear recognition has rich biometric and physiological features, which are stable and do not vary with change in age. Additionally, the facial expression of an individual does not have any impact on the biometry of the ear (Basit, Javed, and Anjum 24- 26). Its large size as compared to the fingers and eye makes it easily recognizable from a distance. Shen, Wang, Wang, Bai, and Zhang point out that with the evolving use of biometric recognition of people, the ear is emerging as a significant source of people recognition. The ear provides more accurate recognition ability when it is used together with the facial recognition devices. Most significant is that the ear is not impacted by any external factors such as makeup, gasses, hair or other artificial settings as such are not able to alter the biometry of the ear (Anika Pflug, Christoph and Busch, 2012). Satish Ravindran (2014) reveals that the ear recognition is sensitive to any external force or factor that interferes with the ordinary operations and working phenomenon of the ear. In particular, it is sensitive to scale variance, invariance in initialization and the level of noise tolerance it can withstand within a particular environment. According to Hurley David (2013) all these factors will adversely affect the ear recognition system depending on their level and magnitude in the environment. The duration of exposure and the condition of the ear will also determine how these factors cause adversaries to the recognition ability. In essence, the ear recognition system will be very sensitive to any changes in the levels of these factors in the environment, and in any case, it finds it difficult to adjust to normalcy, then it will end up losing its recognition ability (Naser Damer & Benedikt Fuhrer, 2012). Care must, therefore, be taken to ensure that all the factors mentioned above are within the limits that the ear can tolerate. In this paper a discussion on the deep contoured ear recognition technique is presented in which normalization, cropping enhancement occur in the early stage. A boundary is then extracted from these images through a deep contour method to be used in the next stage. CHAINLETS is employed to extract the features of the ear and a fast histogram intersection distance used to match the images. In this paper extra attention is given to the deep contour technique to engage the CHAINLEST in extracting the entire ear biometry. The various sizes of the images produced will generate different strictures of the ear biometry which will later be used to classify and match the ear features. Methodology Deep Contoured Method This study borrowed much from the deep contour technique of edge detection to classify all the images obtained for the ear biometry from the background. For the technique, white was used for the ear edge while the black coloration was used for the background of the ear. The deep contour method is a technique that employs