Research Of The Ship Waterline Detection Method Based On Video Information

With the development of transportation, marine transportation attracts more and more attention. Water gauge is a method which is widely used in weighing bulk commodity. The results gained from water gauge can be a powerful proof in commercial disputes. Therefore, the accuracy of draft marks detection can protect the shipper, checker and the consignee's interests. Research on draft marks detection is significant. The accurate detection of waterline will reduce overload phenomenon, which can help to claim ships which make damages to lane and protect the channel resources effectively.The research in this paper is primary on how to get the position of the waterline between the numbers on draft marks and design the automatic detection system of waterline by digital image processing technology. The detection of waterline, segmentation and recognition of the numbers are gained by the algorithm in digital image processing. Digital image processing technology as an convenient method is employed to detect waterline from images which greatly improves efficiency and accuracy of detection. The detecting system is accurate and has few detecting mistakes.Several image processing operations are utilized for the waterline detection in the paper. Compared to other operations, the canny operator gets the better results in detection. In order to find the exact waterline, the false edges are removed by horizontal projection. At last, the horizontal line is fitted by the least square method.In this paper, according to the draft marks, the distance between the waterline and the scale numbers relative to the bottom edge is calculated. the accurate the scale number segmentation is gained by the binary image projection. And Finally, BP neural network is employed to recognize the scale numbers. Then the position of waterline is figured out.According to the several experiments on many images, the effectiveness of method advanced in this paper is proved.