Development of a vision system for ship hull inspection
Loading...
Date
2007
Authors
Zainal Abidin, Zulkifli
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This work introduces a strategy to improve the performance of visual ship hull
inspection using a Remotely-Operated Vehicle (ROV) as its underwater vehicle
platform. The proposed method is aimed at developing a system that will maintain the
camera viewing angle parallel to the ship hull surface. This system consists of four
main units, namely, camera, laser pointers, pan-tilt platform and Cartesian platform.
The position of the camera and the laser pointers are fixed. The system integrates
laser beam tracking and underwater vehicle guidance technique. Initially, ,the region of
interest (ROI), based on laser intensity input, is tracked with dynamic ROI estimation
technique. Then, the feature extraction process will acquire the position of each laser
points in the frames, and calculate the distance of the laser source and the surface.
The range measurement was performed using the modified laser triangulation
technique. This application focuses on binary images with extension of gray level
concept. A closed-loop control system has been developed to classify the camera
positions. An algorithm for overall system control has also been developed, using Intel
Open CV and Visual C++. This system has been applied to the Cartesian platform.
Several experiments using scaled-down ship hull· structure are carried out. The test
results are given and analyze to show its significant result The tracking system is able
to perform horizontal and vertical slices within the range of 17 em and 1 00 em from the
camera to the hull surface and the respond time is about 1 - 3s from the arbitrary
surface to parallel surface on flat surface and 2 - 4s on curve surface .. In short, the
system is found to be very practical and have great potential usefulness for ship hull
inspection technology
Description
Keywords
Vision system , Hull infection