Research On Detection Method Of Ship Draft

With the progress of the society and the rapid development of the economy, shipping is also increasingly strong. In the trading of bulk goods, measuring the ship’s weight is one of the problems which transaction both care about. And in the measurement of the ship’s load, the most widely used method is the draft survey of the ship. Draft survey is on the carrier vessel ship draft, determining the mean draft ship data before and after loading and unloading goods which corresponds to the actual displacement and marine material weight, to calculate the ship load method. It is suitable for situations which have low value or problem to measure instruments in bulk solid weight of the goods, such as coal, ore, chemical fertilizer, etc.The draft system consists of separate parts, which are the image acquisition, handheld terminals, wireless transmission, gap height and line detection subsystem, etc. Ship’s draft data measurement influences the accuracy of final dead weight. The method of ship draft is based on image detection in order to make the more accurate value and intelligent measurement. Water level of high and low affects seawater, which is the characteristic in wave cycle. This thesis uses the climbing robot getting image and adopts image processing method based on projection, image segmentation, color and scale transformation for the design of the detection algorithm.In order to reduce the subsequent processing, process the binarization on the draft image to get the draft scale character high brightness, get high brightness characters individually extraction and rotation correction, match the template to get the character values and location. In order to reduce the influence of scale on both sides of the area of waterline position, intercept water line and the intersection of draft survey area, match the detection result of different color space to finish testing to improve the accuracy of the results. Calculate the qualified results to get the final draft and realize high accuracy.This thesis processes contour detection, tilt correction, and image clipping for ship draft character and the characteristic of the draft image. Through detection of actually collected vessel draft image, it comes out that this method can realize the value of automatic draft detection.