DEVELOPING AN IMAGE PROCESSING FILTER AND VERIFYING IT USING IMAGES TAKEN BY AN AUV IN A TOWING TANK

Underwater Robots such as Remotely Operated Vehicle (ROV) or Autonomous Underwater Vehicle (AUV) and even some submarine are equipped with various cameras so as to inspect underwater environment [1]. Due to the inherent murky condition of underwater, images taken from underwater largely differ from those taken outside water. These differences include limitations in viewing depth, viewing angle, sharpness of image due to suspended particles, perceiving object’s dimensions and colors to name a few [2]. One of the branches in artificial intelligence is about image processing regarding machine vision. In this branch filters could be designed regarding image enhancement, object segmentation, object detection and object recognition. All these filters are devised considering morphology in spatial or frequency domain in series or in parallel or both, and then implemented on target images [3].