The traditional Hausdorff distance only considers the position information when calculating the similarity of ship trajectory, and rarely considers the characteristic information such as ship heading and speed. Secondly, the traditional Hausdorff distance is easily affected by the loss of track points when measuring ship trajectory, resulting in the increase of the distance between ship trajectories. In order to measure the distance between ship trajectories more accurately, this paper proposed a trajectory similarity measurement method based on improved Hausdorff distance. The algorithm took two important behavioral characteristics of ship heading and speed into account. It makes the distance measurement between ship trajectories more accurate and solves the problem that the distance between ship trajectories increases due to the loss of ship trajectory points in the traditional Hausdorff distance measurement. The ship trajectory data from Shanghai South Channel are used in the experiment. Compared with the traditional hausdorff distance, the results show that the clustering results of the proposed method can effectively distinguish the differences in the behavior characteristics of the trajectory
In order to further respond to the national policy and improve the monitoring capability of the sealing status of inland bulk carriers on the Yangtze River, this paper proposes the C-YOLOv5 model on the basis of the YOLOv5 algorithm. The elkan K-Means clustering algorithm is introduced to optimize the target candidate frame so that it can adapt to the detection environment of small targets while ensuring the recall and accuracy of recognition. In order to improve the focus on the detected targets, the SE focus mechanism module is introduced in the head of the model. In addition, some of the ordinary convolutions in the structure are replaced with depth-separable convolutions to further improve the detection speed. The experimental results show that the detection accuracy and speed are improved by using the C-YOLOv5 model, with an accuracy as high as 91.1% and a speed as high as 66.5 f/s. The detection accuracy and retrieval rate of the C-YOLOv5 algorithm both reach 90%. The C-YOLOv5 algorithm's ability to recognize the target state is also significantly improved under poor line-of-sight and dark light test conditions.
To solve the problems of low recognition accuracy and slow detection of crew fatigue driving behavior in the cockpit of ships during the process of sailing in and out of the port, the SSD model was studied. By replacing its backbone network and improving the prior frame generation mechanism, the MV-SSD model is proposed. Replace the backbone network VGG16 in the original SSD model with MobileNetV3, reducing the network parameters of the backbone network. Using the K-means algorithm to cluster the real detection boxes in the face area dataset, the prior box allocation mechanism of the SSD model was redesigned, reducing the number of prior box generation by nearly half, and the ERT algorithm in the Dlib library is combined to locate the face key points, and finally, the PERCLOS criterion is used to determine whether the driver is fatigued. Experimental results show that the average accuracy (mAP value) of the MV-SSD model is 7.15% higher than that of the original SSD model, and the detection speed (FPS value) is increased by 98frames/s, which is more suitable for the detection of the crew face area, and the average accuracy of the constructed fatigue detection algorithm for fatigue features is more than 94%.
An investigation was conducted to predict ship trajectories and discern their navigational intent in intersecting waters. The study involved experiments utilizing AIS data near the Baoshan District, Shanghai, and the development of the RBFNN+KNN model. The data preprocessing steps encompassed: 1) cleaning abnormal trajectory points and routes; 2) employing the LOF algorithm to filter trajectories with significant outliers; 3) clustering ship trajectories. Two clusters, representing routes in the straight-ahead channel and the right-turn channel, were selected as the sample dataset. The RBF neural network was employed to forecast ships' trajectories in future time periods, and the KNN classification algorithm was integrated to determine the vessels' sailing intentions. The experimental outcomes demonstrated that the proposed method achieved 99.88% accuracy in navigational intent recognition for the training set and 99.18% for the test set. These results indicate the model's superior ability to predict ships' navigational intent.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.