Johannes Wagner finished his Bachelor thesis “Influence of Image Compression on Ear Biometrics” after working with da/sec and CASED in the period of December 2013 to April 2014.
Human ears have been used as major feature in the forensic science in recent years. Due to the fact that ear images can be easily taken from distance without knowledge of the examined person, the ear is suitable for security and surveillance use cases.
Most surveillance cameras lack in decent image quality for usage with biometric recoginiton algorithms. While actual international biometric standards don’t consider requirements and guidelines for acquiring ear data, the aim of this thesis is, to build different scenarios with ear image databases for evaluating the image compression impact with state-of-the-art compression standards versus uncompressed image data.
The thesis put focus on the optimal choice of compression algorithms and rates for biometric ear images for utilization with ear detection and recognition algorithms.