Range Gated Imaging System for Underwater Robotic Vehicle

Abstract

This paper introduces the work that has been done to improve the underwater visibility from underwater robotic vehicle (URV) for sub-sea inspection and repair missions. It is undertaken as part of a research program to develop an underwater imaging system in turbid water condition. Images associated with underwater imaging systems are frequently degraded due to absorption and scattering effects from its underwater environment. The absorption effect reduces the signal strength, and the latter effect reduces both signal strength and image resolution. In order to improve the underwater visibility in turbid conditions, general design considerations of the imaging system are planned in two stages, namely: hardware upgrading and system optimization. We demonstrate this concept by improving the underwater visibility with an advanced technique -range gated imaging system (hardware upgrading), and the optimization stage in terms of tail-gating and image processing techniques. Tail-gating is realized by a delay in camera gating towards the tail of reflected image temporal profile (RITP). It is followed by contrast limited adaptive histogram equalization (CLAHE) to further enhance the range-gated images. A quantitative image quality measure, modified fidelity index (MF), is used to evaluate the enhanced imaging techniques. The measure is based on 2 dimensional grayscale images of USAF (United State Air Force) targets. These targets have been used extensively in underwater imaging system development. It has resolution bars in various frequencies and arrangement, which enable spatial frequencies and signal strength analysis. In the first stage, the MF index shows at least 40% improvement from non-gated to gated images in increased turbidity. By comparing images between Front-gated to the Tail-gated RITP quantitatively, the latter is improved by about 4% -22% (based on MF index) in various turbidity levels. Finally, the CLAHE technique further improves the gated images in terms of contrast gain.