5 September 2022 in IEEE Transactions on Geoscience and Remote Sensing
Object tracking in satellite videos faces various challenges such as target occlusion, target rotation, and background clutter. This study proposes a correlation particle filter algorithm with motion estimation for object tracking in satellite videos. The tracker, called CPKF, combines the strengths of the correlation, particle, and Kalman filters. Compared with existing tracking methods based on correlation filters, the proposed tracker has three major advantages: (1) Particle sampling, and motion estimation build robustness against partial and complete occlusion. (2) Color histogram model makes it robust to target rotation. (3) Fusion of multiple feature response maps effectively handle background clutter and low contrast. The experimental results demonstrate that the proposed tracking algorithm performs better than state-of-the-art methods.
10 January 2023 in IEEE Transactions on Geoscience and Remote Sensing
Fine-grained ship classification has many applications in military and civilian fields. In recent years, deep learning has been widely used for classification tasks, and its success is inseparable from that of big data. However, ship images are valuable, with only a few images of a specific category being obtained, leading to the fine-grained few-shot ship classification problem. In addition, feature map channels contain distinct characteristics and discriminative details, which significantly influence fine-grained ship classification. Intuitively, channels with distinct characteristics should be assigned larger weights for classification, but most few-shot learning methods treat the channels equally. Therefore, we propose a generalized ridge regression-based channel-wise feature map weighted reconstruction network to address these issues. First, we reconstruct the query feature map by assigning different weights to the support feature map channels using the generalized ridge regression method. The channels with large discriminative details contribute more towards reconstruction. Second, we propose a support channel weight module to calculate the channel weight matrix used in the generalized ridge regression method. Finally, based on the reconstructed query feature map, we can calculate the reconstruction error. The reconstruction error is adopted as the distance metric. Our proposed method achieves excellent performance on the fine-grained ship, bird, aircraft and WHU-RS19 datasets compared to other representative few-shot learning methods. Considering the limited studies on the fine-grained few-shot ship classification problem, we believe that our work is of great significance.
17 March 2023 in IEEE Geoscience and Remote Sensing Letters
Recent deep learning based-works have made remarkable progress in Remote-Sensing Image Super-Resolution (RSISR). However, the complicated network architecture as well as a huge amount of parameters increase computational cost, hindering their practical deployment. To alleviate this problem, we propose a novel Re-parameterized Feature Distillation Network (ReFDN) for lightweight and efficient RSISR tasks. Feature distillation, refinement, condensation, and enhancement are efficiently integrated into the re-parameterized feature distillation block named ReFDB for lighter and stronger feature extraction. With the help of elaborate re-parameterized convolution (ReConv) design, we further boost the feature refinement capability without extra inference costs. Additionally, we design an efficient channel and spatial attention module (ECSA) to enhance the important objects and regions of the intermediate features adaptively. Conducted on both commonly used datasets and additional Google Earth data, the experimental results demonstrate our method can achieve a good trade-off between SR performance and network complexity. Our code will be publicly available at https://github.com/DaxingZ/ReFDN.
30 December 2022 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Ship detection with several military and civilian applications has drawn considerable attention in recent years. In remote sensing images, ships have the characteristics of arbitrary orientation. Based on the characteristics, many arbitrary-oriented ship detectors have been proposed. Most of these detectors preset many horizontal or rotated anchors and determine the positive and negative samples based on the intersection-over-union (IoU) between the anchor and ground-truth bounding box, in what is called the label assignment process. However, IoU performance is limited as it can only reflect the quality of the anchor to a certain extent. In addition, the manually fixed IoU threshold to separate the positive and negative limits the flexibility of the method, as different ships may have different optimal thresholds. Moreover, the equally weighted training samples cause a misalignment between the classification and regression heads. Therefore, we propose a dynamic soft label assignment method for arbitrary-oriented ship detection. First, we design a novel anchor quality score function that takes into account both prior and prediction information of the anchor and enables the model to participate in the label assignment process. Second, we propose a dynamic anchor quality score threshold instead of a fixed IoU threshold for dividing positive and negative samples. Third, in contrast to assigning equal weights, we propose a soft label assignment strategy to weigh the training samples in the loss function. The proposed method offers superior detection performances for arbitrary-oriented ships with only one horizontal preset anchor. Experimental results on HRSC2016, FGSD, and ShipRSImageNet datasets demonstrate the effectiveness of our proposed dynamic soft label assignment for arbitrary-oriented ship detection.
25 May 2022 in IEEE Geoscience and Remote Sensing Letters
Interpreting airborne remote sensing images plays an important role in aviation control and battlefield situational awareness. However, highly dynamic aircraft detection remains challenging, owing to variable object sizes, flexible attitudes, and motion blur. This study develops a multi-scale airborne aircraft dataset benchmark to overcome aircraft detection challenges, such as high intraclass variance, multiple scales and angles, motion blur, and partial occlusion. We also propose a trained-from-scratch aircraft detector, the local feature correlation single shot multibox detector (LFC-SSD), to detect multi-scale aircraft. The LFC-SSD comprises a local correlation feature extraction module, called “Right-Residual,” and a feature fusion module using a reverse feature pyramid network (R-FPN). Right-Residual extends the global receptive field by aggregating contextual information while learning non-adjacent region features efficiently. R-FPN utilizes multi-path information transfer horizontally with recursive integration to enhance the robust representation of the location information of multi-scale object features. In addition, a specific default boxes method is designed for remote-sensing images of aircraft. Extensive experimental results confirm the significant improvement of the proposed method over several existing state-of-the-art methods.
19 May 2025 in IEEE Geoscience and Remote Sensing Letters
Fine-grained aircraft classification in optical remote sensing images relies on subtle difference of local parts. However, aircrafts in low-resolution (LR) images are depicted as small objects with blurred details and limited appearance information, resulting in the dramatic classification performance drop. We propose a novel dynamic hybrid knowledge distillation (DHKD) method to enhance fine-grained aircraft classification in LR remote sensing images. The core idea is to construct a teacher network for high-resolution (HR) images and a student network for LR images, and dynamically distill the key discriminative region attention-guided feature-based and logit-based object prior knowledge acquired from the teacher to the student to instruct its training, thereby improving the student’s fine-grained representation capability and classification performance for LR images. The distillation weights of different knowledge types are adaptive to their optimization direction similarity with that of the primary fine-grained classification task during the iterations. Extensive experiments on the MAR20 dataset demonstrate that DHKD consistently outperforms state-of-the-art distillation methods on fine-grained classification for aircrafts with absolute sizes of 32, 20, and 12 pixels in LR remote sensing images.