Simulation Study And Analysis On Errors Of Ice Thickness Estimations Using Ship-board Digital Camera Based On Laboratory Tests

With the development of digital technology, the digital camera is currently using in capturing the broken sea ice cross section. And the method has been used in polar exploration widely, as well as in monitoring the sea ice based on the offshore platform. We can obtain ice thickness from such images by image processing technology. But, whether in the route or the offshore platform, it is difficult and dangerous to have field survey on the ice surface. Then it is impossible to compare the sea ice thickness captured by digital camera from the true value. In order to study the technique in monitoring sea ice by ship-based CCD camera, we conducted a laboratory physical model experiment, and further elaborate the errors. In addition, we compare the dates captured by single CCD from the stereo imaging system consisted of a pairs of CCD.Firstly, this paper analyses the impacts of the three-dimensional angles resulted from the sea ice's roll and CCD focusing distance on observation errors in detail. We describe the sea ice's roll beside ship side as rigid motion in the space right angle coordinate systems. The angles of sea ice rotate about the X,Y,Z axes are defined asα,β,γ. And yet the camera focusing distance can not be uniform parameters. So we translate the focusing distance into the images'width proportions. And then we analyses the impacts ofα,β,γand the images' width proportions on observation errors by linear regression and surface fitting methods.Secondly, the paper analyses the human errors in detail. We select the same sea ice model images which are level and parallel to the ship's rail (α=β=γ=0°). It can eliminate the effects of sea ice model's roll. And then we get the difference of the human errors through five peoples process the same sea ice model image.Finally, we compare the dates captured by single CCD from the stereo imaging system. It is found that when the sea ice model images are level and parallel to the ship's rail (α=β=γ=0°), the best observational data captured by single CCD is better than the observational data captured by the stereo imaging system consisted of a pairs of CCD. But it is the most exceptional circumstances. In most cases it is not exist. The observational data captured by the stereo imaging system consisted of a pairs of CCD is better, so the author advise that it is a significant research field to applicant this technique on Sea ice observation.