Multi-object tracking in satellite videos has been widely used in civilian and military fields. Among them, the tracking of vehicles has important applications in the field of traffic monitoring. However, the tracking of vehicles in satellite videos still remains challenging and unsolved due to the extremely small size and the lack of appearance and geometric features. In this paper, we propose an improved SORT to tackle the tracking of vehicles in satellite videos by introducing C3D to CenterNet to improve the detection performance and promote the overall tracking performance. Specifically, we use C3D as the backbone of CenterNet to extract spatio-temporal information and use a 3D channel attention mechanism to fuse the information extracted by C3D to improve the detection performance, thereby improving the tracking results. The qualitative and quantitative results of experiments on videos of Jilin-1 satellite constellation show that our method can efficiently improve the tracking performance of vehicles in satellite videos.
Benefiting from the superiority visual capacity of infrared imaging system such as long operating distance and all-weather work, infrared target detection is widely applied in weapon guidance, security protection and other fields. Aiming at the characteristic of infrared targets, we propose an infrared detection algorithm based on feature fusion. Firstly, gaussian filter and background suppression are employed to improve signal-to-noise ratio and increase the contrast between target and background, respectively. Secondly, a feature fusion architecture is added to backbone for feature extraction, so as to obtain features with rich information. Finally, in order to solve the problem of imbalance between positive and negative samples, the network is trained by focal loss to improve training performance. We evaluate the proposed detection algorithm on public infrared image datasets. The results demonstrate that the proposed algorithm can achieve promising performance.
Infrared target tracking is a fundamental task and plays an important role in military, weapon guidance and other fields because of the superiority of infrared imaging system. However, it is a challenging task due to lower resolution, smaller size and less texture information of infrared targets. Although varieties of methods demonstrate promising performance on visual target tracking, they cannot directly perform well on infrared images. To address this issue, we propose an infrared target tracker based on siamese network, named InfTrans, consisting of three components: a feature extraction backbone, an encoder-decoder transformer and a prediction head. In the InfTrans, transformer architecture is employed to exploit spatial and temporal information effectively. Firstly, considering the superior ability in capturing long-range dependencies, encoder-decoder transformer is connected after backbone to capture relationship between template and search patch along spatial and temporal dimension. Secondly, considering the parallel computing power of transformer, InfTrans receives search patch sequences containing multiple patches as input of search branch and processes multiple frames in parallel. Finally, a prediction head is employed to output bounding box sequences. InfTrans views infrared target tracking as a parallel bounding box prediction problem. Extensive experiments show that the proposed tracker achieves promising performance on public infrared dataset.
Image restoration is a necessary preprocessing step for infrared remote sensing applications. Traditional methods allow us to remove the noise but penalize too much the gradients corresponding to edges. Image restoration techniques based on variational approaches can solve this over-smoothing problem for the merits of their well-defined mathematical modeling of the restore procedure. The total variation (TV) of infrared image is introduced as a L1 regularization term added to the objective energy functional. It converts the restoration process to an optimization problem of functional involving a fidelity term to the image data plus a regularization term. Infrared image restoration technology with TV-L1 model exploits the remote sensing data obtained sufficiently and preserves information at edges caused by clouds. Numerical implementation algorithm is presented in detail. Analysis indicates that the structure of this algorithm can be easily implemented in parallelization. Therefore a parallel implementation of the TV-L1 filter based on multicore architecture with shared memory is proposed for infrared real-time remote sensing systems. Massive computation of image data is performed in parallel by cooperating threads running simultaneously on multiple cores. Several groups of synthetic infrared image data are used to validate the feasibility and effectiveness of the proposed parallel algorithm. Quantitative analysis of measuring the restored image quality compared to input image is presented. Experiment results show that the TV-L1 filter can restore the varying background image reasonably, and that its performance can achieve the requirement of real-time image processing.
As one of the most significant methods to study laser propelled rocket, the numerical simulation of laser propulsion has drawn an ever increasing attention at present. Nevertheless, the traditional serial simulation model cannot satisfy the practical needs because of insatiable memory overhead and considerable computation time. In order to solve this problem, we study on a general algorithm for laser propulsion design, and bring about parallelization by using a twolevel hybrid parallel programming model. The total computing domain is decomposed into distributed data spaces, and each partition is assigned to a MPI process. A single step of computation operates in the inter loop level, where a compiler directive is used to split MPI process into several OpenMP threads. Finally, parallel efficiency of hybrid program about two typical configurations on a China-made supercomputer with 4 to 256 cores is compared with pure MPI program. And, the hybrid program exhibits better performance than the pure MPI program on the whole, roughly as expected. The result indicates that our hybrid parallel approach is effective and practical in large-scale numerical simulation of laser propulsion.
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.