Automatic identification systems (AIS) can record a large amount of navigation information about ships, including abnormal or illegal ship movement information, which plays an important role in ship supervision. To distinguish the trajectories of ships and analyze the behavior of ships, this paper adopts the method of supervised learning to classify the trajectories of ships. First, the AIS data for the ships were marked and divided into five types of ship tracks. The Tsfresh module was then used to extract various ship trajectory features, and a new ensemble classifier based on traditional classification using a machine learning algorithm was proposed for modeling and learning. Moreover, ten-fold cross validation was used to compare the ship trajectory classification results. The classification performance of the ensemble classifier was better than that of the other single classifiers. The average F1 score was 0.817. The results show that the newly proposed method and the new ensemble classifier have good classification effects on ship trajectories.
1 November 2023 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Remote-sensing ship monitoring is a crucial area of research with key applications in military and civilian fields. The ability to extract information, such as ship length, width, and heading from remote-sensing data, particularly from synthetic aperture radar (SAR) images, is of paramount importance. Current state-of-the-art SAR image ship monitoring focuses primarily on ship detection. Assessing the direction of ships usually relies on the observability of wake features. However, the observability of these wake features is often affected by factors, such as the SAR system parameters, ship attributes, and dynamic marine environments. This can make accurate direction assessments a challenging task. In response to these challenges, this study has presented a novel and effective algorithm for ship monitoring from SAR images based on an anchor-free framework and the powerful feature extraction capabilities of convolutional neural networks. The proposed method learned the scattering and morphological information of a ship's bow and stern from high-resolution SAR images to determine the ship's direction with a high level of accuracy using a rotating bounding box. The algorithm was tested on a dataset, achieving an average precision of 90.8% and bow classification accuracy of 92.5%, demonstrating its potential contributing to the advancement of remote sensing.
15 May 2024 in IEEE Geoscience and Remote Sensing Letters
Ship target search tasks aim to match specific ships across two or more satellite images. Like pedestrian and vehicle re-identification tasks in computer vision, accurate ship re-identification encounters challenges, including subtle differences between ships of the same type and substantial intra-instance variations due to satellite angle of view and spectral differences. To tackle these challenges, this paper introduces a deep learning-based two-branch framework for ship target search, integrating ship detection and re-identification tasks. One branch extracts the target ship features while the other captures the search region features. These features are then fused through a dedicated layer, and the final output is derived from the keypoint detection header. A new dataset was curated using Sentinel-2 and Gaofen-1 satellite data. Experimental results validate the robustness of the proposed method, achieving an accuracy of 94.37% on the new dataset. Our method’s scalability has been validated through experiments using CBERS-04 and Gaofen-6 satellite data.
This paper proposes a new model-based gait recognition method. Different from other methods using 3D (3-dimensional) keypoint information and skeleton information, we directly stack the 2D (2-dimensional) keypoint heatmaps in the gait sequence in the time dimension, and input it into the network structure based on 3D-CNN (3-dimensional-convolutional neural network). Then, through the gait analysis on the two dimensions of time and space, the effective gait features are finally obtained. Compared with other model-based methods, this method is more clear, concise and elegant in the process of feature extraction. The test of CASIA-B dataset shows that in the model-based gait recognition method, we have competitive performance.
6 April 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Ship motion analysis using remote sensing imagery is essential for maritime surveillance, traffic monitoring, and ocean management. Existing optical detection methods, however, have primarily concentrated on identifying static ship targets through horizontal or rotated bounding boxes, and therefore cannot effectively estimate dynamic attributes such as heading and velocity. In this work, we develop DPMVectorNet (Dual-Point Motion Vector Network), an end-to-end learning framework designed for the joint estimation of ship position, heading, and speed from medium-resolution multi-temporal satellite imagery. Specifically, the framework is evaluated on multi-temporal ASTER VNIR data (15 m spatial resolution), in which temporally adjacent observations are combined to construct RGB composite images, which are then processed by a convolutional neural network to capture and fuse multi-scale spatial features. Ship motion is represented through the temporal matching of two bow keypoints, from which the movement direction is derived, while the velocity is computed from their displacement and the corresponding imaging interval. Experiments conducted on multi-temporal ASTER data show that DPMVectorNet achieves an F1-score above 90% for ship detection, with a mean heading error below 2°, and a velocity deviation within 1 m/s when compared with Automatic Identification System (AIS) records. These results demonstrate that the proposed framework enables accurate ship detection and reliable heading and velocity estimation from medium-resolution imagery by effectively exploiting multi-temporal information.