There
is a growing interest in underwater applications. Stereo vision is one of the
best methods for distance estimation of underwater object. In this research two
pairs of cameras were used as stereo image acquisition to estimate the distance
of underwater object. The stereo vision system in this project consists of
calibration of camera, rectification of images, segmentation of images, finding
of centroid and localization of object. Edge-based segmentation, Mathematical
morphology and largest area selection are used to perform image segmentation.
Finally, It will be shown that curve fitting is better than triangulation
method to estimate the coordinates with the overall error of around 0.5 cm with
water condition where the overall error of using triangulation method is around
2.2 cm which is too much in range estimation
Website: http://www.arjonline.org/engineering/american-research-journal-of-mechanical-and-automation-engineering/
Website: http://www.arjonline.org/engineering/american-research-journal-of-mechanical-and-automation-engineering/
No comments:
Post a Comment