Ship waterline extraction strategy based on deep learning

A method for detecting the position of ship draught based on deep neural network is proposed. Compared with the traditionally selected features, the features learned by the deep neural network-based method have strong robustness and stability, and can adapt to new objects that have not appeared in the training set. The method firstly uses the semantic segmentation algorithm based on deep learning to segment the target region in the image, and obtains the position of the waterline in the image through horizontal projection, and then obtains the final draft depth according to the statistical method. Experiments show that the proposed method can accurately segment the target region in the image, and then calculate the waterline value. By comparing with the results obtained by manual, and the results are proved to be effective. © 2020, Editorial Department of Control Theory & Applications South China University of Technology. All right reserved.