Cubic spline regression based enhancement of side-scan sonar imagery

摘要

Exploring the seas and the oceans is essential for industrial and environmental applications. Given the fact that the seas cover 72% of the surface of the Earth and are home to 90% of all life found on it, underwater imaging has become an active research area in recent years. Due to the high absorption of electromagnetic waves by water, sonar is currently the exemplary choice used in underwater imaging. Yet, underwater images acquired with sonars suffer from various degradations, since the sound signal is affected by the environment and the sonar parameters and geometry. This work proposes an enhancement method that aims at getting close to natural underwater images. The enhanced images can be used in further applications related to seabed mapping and underwater computer vision. The enhancement aims at reducing the echo-decay and some effects of the receiver gain.