A remote sensing image target recognition method based on improved Mask-RCNN model

Yu, Huan; Xu, Fuxin

doi:10.1109/icbaie52039.2021.9389916

Cited by 7 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The spatially oriented object detection framework that Yu et al [59] developed had the highest accuracy and resilience in remote sensing images and showed significant application potential in remote sensing image analysis. While this was going on, the upgraded Mask-RCNN model [60] was used to recognize objects in remote sensing images, and the segmentation accuracy and precision were enhanced using new techniques for multi-scale feature fusion, regiongenerating networks, and segmentation precision optimization. In addition, the aircraft detection method based on multi-scale Faster-RCNN successfully dealt with complex scenes and images of different scales by introducing multi-scale feature pyramids and RoI (region of interest, RoI) complete utilization strategy [61] and achieved excellent detection performance.…”

Section: Remote Sensing Object Detection Algorithm Based On Candidate...mentioning

confidence: 99%

A Review: Remote Sensing Image Object Detection Algorithm Based on Deep Learning

Bai,

2023

Electronics

View full text Add to dashboard Cite

Target detection in optical remote sensing images using deep-learning technologies has a wide range of applications in urban building detection, road extraction, crop monitoring, and forest fire monitoring, which provides strong support for environmental monitoring, urban planning, and agricultural management. This paper reviews the research progress of the YOLO series, SSD series, candidate region series, and Transformer algorithm. It summarizes the object detection algorithms based on standard improvement methods such as supervision, attention mechanism, and multi-scale. The performance of different algorithms is also compared and analyzed with the common remote sensing image data sets. Finally, future research challenges, improvement directions, and issues of concern are prospected, which provides valuable ideas for subsequent related research.

show abstract

Section: Remote Sensing Object Detection Algorithm Based On Candidate...mentioning

confidence: 99%

A Review: Remote Sensing Image Object Detection Algorithm Based on Deep Learning

Bai,

2023

Electronics

View full text Add to dashboard Cite

show abstract

“…On the other hand, there is an increase in the type and number of remote sensing data (Cheng, Zhou, and Han 2016;Xia et al 2018;Zhang et al 2019;Li et al 2020c), which is collected from various satellites, such as Google Earth, Gaofen series, and Jilin series. Current works (Zhao et al 2022c,a;Huiming and Fuxin 2021;Cui et al 2020;Sumbul, Cinbis, and Aksoy 2019;Fang et al 2021;Li et al 2020d) adopt convolutional neural networks (CNNs) to improve the performance of RSOR, where the training and testing data manifest similar distributions. However, RSOR in real applications are varied from Figure 1: Generalization comparison among ours, MoEx (Li et al 2021a) and NTS (Yang et al 2018).…”

Section: Introductionmentioning

confidence: 99%

Style-Content Metric Learning for Multidomain Remote Sensing Object Recognition

Zhao

Yang

Liu

et al. 2023

AAAI

View full text Add to dashboard Cite

Previous remote sensing recognition approaches predominantly perform well on the training-testing dataset. However, due to large style discrepancies not only among multidomain datasets but also within a single domain, they suffer from obvious performance degradation when applied to unseen domains. In this paper, we propose a style-content metric learning framework to address the generalizable remote sensing object recognition issue. Specifically, we firstly design an inter-class dispersion metric to encourage the model to make decision based on content rather than the style, which is achieved by dispersing predictions generated from the contents of both positive sample and negative sample and the style of input image. Secondly, we propose an intra-class compactness metric to force the model to be less style-biased by compacting classifier's predictions from the content of input image and the styles of positive sample and negative sample. Lastly, we design an intra-class interaction metric to improve model's recognition accuracy by pulling in classifier's predictions obtained from the input image and positive sample. Extensive experiments on four datasets show that our style-content metric learning achieves superior generalization performance against the state-of-the-art competitors. Code and model are available at: https://github.com/wdzhao123/TSCM.

show abstract

MIMO Radar Waveform Joint Optimization in Spatial-Spectral Domain for Anti-Interference

Chu

Zhang²,

Chen

2021

IEEE Access

View full text Add to dashboard Cite

Aiming at the target detection problem of MIMO radar under the condition of both spatial and spectral domain interferences, a waveform design method for MIMO radar based on joint optimization of spatial and spectral domain interferences suppression in multi-target scenarios is proposed. Compared with the traditional synthetic waveform method based on the transmitting signal covariance matrix, this method directly performs waveform design in the space-spectral domain in turn, considering target characteristics and spatial-spectral environment comprehensively. Firstly, according to the target characteristics and energy allocation strategy, the fast time transmission pulse set is divided into a group of sub-pulse sets whose number equals to the number of targets. Meanwhile a set of transmit filter weighting factor vectors are introduced and the number of vectors equals to the number of spatial interferences. Furthermore, the spatial waveform optimization model is constructed according to the pattern matching criterion, subjecting to the peak-to-average power ratio of the transmitted signal. The sub-pulse sets are obtained through convex optimization, and the full matrix of the emission waveform is formed by the subpulse sets. Then, according to the invariability of radar transmit pattern with changing the initial phase of the sub-pulse, the transmitted waveform with desired beam pattern and spectrum distribution is obtained by optimizing the initial phase matrix through convex optimization. Finally, the optimized transmit waveform of MIMO radar with the ability of simultaneously suppressing space frequency interferences is generated. The simulation results prove the effectiveness of the proposed method.INDEX TERMS MIMO radar; beam pattern matching; spatial-spectral domain anti-jamming; waveform synthesis

show abstract

A remote sensing image target recognition method based on improved Mask-RCNN model

Cited by 7 publications

References 6 publications

A Review: Remote Sensing Image Object Detection Algorithm Based on Deep Learning

A Review: Remote Sensing Image Object Detection Algorithm Based on Deep Learning

Style-Content Metric Learning for Multidomain Remote Sensing Object Recognition

MIMO Radar Waveform Joint Optimization in Spatial-Spectral Domain for Anti-Interference

Contact Info

Product

Resources

About