Dam Structure Deformation Monitoring by GB-InSAR Approach

Qiu, Zhiwei; Ming-lian, Jiao; Jiang, Tinchen; Zhou, Li

doi:10.1109/access.2020.3005343

Cited by 29 publications

(14 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such images, in combination with machine learning algorithms, enable surface deformation monitoring [2][3][4][5][6], snow avalanche identification [7], bridge [8] and dam structure monitoring [9], open pit mine safety management [10], and terrain classification [11]. Moreover, machine learning enables object classification using only radar images.…”

Section: Introductionmentioning

confidence: 99%

Deep Learning Approach for Object Classification on Raw and Reconstructed GBSAR Data

et al. 2022

View full text Add to dashboard Cite

The availability of low-cost microwave components today enables the development of various high-frequency sensors and radars, including Ground-based Synthetic Aperture Radar (GBSAR) systems. Similar to optical images, radar images generated by applying a reconstruction algorithm on raw GBSAR data can also be used in object classification. The reconstruction algorithm provides an interpretable representation of the observed scene, but may also negatively influence the integrity of obtained raw data due to applied approximations. In order to quantify this effect, we compare the results of a conventional computer vision architecture, ResNet18, trained on reconstructed images versus one trained on raw data. In this process, we focus on the task of multi-label classification and describe the crucial architectural modifications that are necessary to process raw data successfully. The experiments are performed on a novel multi-object dataset RealSAR obtained using a newly developed 24 GHz (GBSAR) system where the radar images in the dataset are reconstructed using the Omega-k algorithm applied to raw data. Experimental results show that the model trained on raw data consistently outperforms the image-based model. We provide a thorough analysis of both approaches across hyperparameters related to model pretraining and the size of the training dataset. This, in conclusion, shows how processing raw data provides overall better classification accuracy, it is inherently faster since there is no need for image reconstruction and it is therefore useful tool in industrial GBSAR applications where processing speed is critical.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep Learning Approach for Object Classification on Raw and Reconstructed GBSAR Data

et al. 2022

View full text Add to dashboard Cite

show abstract

“…[1][2] In a complex and changeable marine environment, the safety obstacle avoidance technology of AUV is not only a guarantee of its navigation and work but also an important part of its navigation control technology. With the deepening of ocean exploration in various countries, further improvement of the dynamic obstacle avoidance capability of AUVs in complex marine environments has become the key to restricting the effectiveness of AUVs [3][4] .…”

Section: Introductionmentioning

confidence: 99%

AUV Dynamic Obstacle Avoidance Method Based on Improved PPO Algorithm

Zhu¹,

Zhou²,

Liu³

et al. 2022

IEEE Access

View full text Add to dashboard Cite

A reasonable obstacle avoidance method for AUV 3D path planning is difficult. Existing obstacle avoidance methods have certain drawbacks. For example, they are only applicable to 2D planar applications and cannot effectively handle dynamic obstacles. To address these problems, this study designs an obstacle collision prediction model (CPM). Based on the results of the simulation operation of the obstacle inertial motion, the safety of the AUV navigation is evaluated to improve the sensitivity of the model to dynamic obstacles. Then, the learning ability of the sequence sample data is enhanced by combining it with the long short-term memory (LSTM) network, and the training efficiency and effect of the algorithm are improved. The trained proximal policy optimization (PPO) network can output reasonable actions to control the AUV to avoid obstacles, forming an AUV 3D dynamic obstacle avoidance strategy based on CPM-LSTM-PPO algorithm. The simulation results show that the proposed algorithm has good generalization in uncertain environments, and successfully realizes dynamic obstacle avoidance of AUV in different three-dimensional unknown environments, providing theoretical and technical support for real path planning.INDEX TERMS AUV; dynamic obstacle avoidance; deep reinforcement learning; proximal policy optimization algorithm; collision prediction model.

show abstract

“…The development of devices that aid ground surface disaster prevention is in great demand worldwide, particularly in countries that experience unstable environmental changes such as earthquakes, rainfall, and severe weather. Ground-based synthetic aperture radar (GB-SAR) has been used effectively to measure ground surface displacements [1][2][3][4][5]. A conventional GB-SAR system comprises transmitting and receiving horn antennas sliding on a metal bar to cover the required synthetic aperture.…”

Section: Introductionmentioning

confidence: 99%

“…Two λg/2 open circuit stubs are added to the array to reflect the power traveling toward the ends. The amplitude distribution at each array element is extracted using (1).…”

Section: Introductionmentioning

confidence: 99%

Antenna Array for Ku-Band MIMO GB-SAR

El-Hameed

Suzuki

2021

IEEE Access

View full text Add to dashboard Cite

A new multi-input multi-output (MIMO) antenna configuration with wideband and low sidelobe level (SLL) is introduced for ground-based synthetic aperture radar (GB-SAR). The proposed MIMO antenna has 15 transmitting and 16 receiving antenna arrays, arranged in two parallel straight lines, to synthesize a virtual array of 256 elements suitable for monitoring relatively large areas. The inter-element spaces are optimized to λ between transmitting elements and 4λ between receiving elements, to realize both low mutual coupling and an inferior SLL of lower than-37 dB for the overall MIMO antenna. The MIMO antenna element has eight rectangular patch linear arrays working at 17.1 GHz. A traveling wave-feeding network with center excitation is used to feed each element, in order to reduce the SLL and achieve a wide bandwidth of more than 300 MHz. The antenna array shows a gain of 14.3 dBi and half-power beamwidths (HPBWs) of 10° and 105° in the elevation and azimuth planes, respectively. The performance of the proposed design satisfies the GB-SAR requirements, which are validated by fabrication and measurement.

show abstract

Dam Structure Deformation Monitoring by GB-InSAR Approach

Cited by 29 publications

References 14 publications

Deep Learning Approach for Object Classification on Raw and Reconstructed GBSAR Data

Deep Learning Approach for Object Classification on Raw and Reconstructed GBSAR Data

AUV Dynamic Obstacle Avoidance Method Based on Improved PPO Algorithm

Antenna Array for Ku-Band MIMO GB-SAR

Contact Info

Product

Resources

About