Search

Your search keyword '"Zhai, Jingsheng"' showing total 146 results
Start Over Author "Zhai, Jingsheng"
146 results on '"Zhai, Jingsheng"'

10. PID2Net: A Neural Network for Joint Underwater Polarimetric Images Descattering and Denoising

Author

Liu, Hedong, Zhang, Wenjie, Han, Yilin, Li, Xiaobo, Liu, Tiegen, Zhai, Jingsheng, Mao, Yefei, Xiao, Lin, and Hu, Haofeng

Abstract

Underwater optical imaging is widely used in various marine exploration tasks. However, the scattering of light by suspended particles in water will result in poor contrast to underwater images. Underwater polarimetric imaging could effectively suppress the influence of scattered light. However, apart from scattering, the absorption and poor illumination will cause low signal to radio of images, and thus the influence of noise on image quality is also significant. Therefore, underwater images should jointly solve scattering and noise problems. In this article, we propose a deep learning-based image enhancement method, called polarimetric image descattering and denoising neural network (PID2Net), to suppress the negative impact of both scattering and noise on underwater image quality. In particular, our network is based on frequency decomposition, which aims to remove scattered light at low frequency and suppress noise at high frequency by applying residual dense network. Experimental results show that our method can well enhance underwater images obtained in dense scattering media and severe noise environments, with performance superior to previous methods.

Published
2024
Full Text
View/download PDF

11. Preliminary sea ice detection results from GNSS-R payload on board Chinese Jilin-1 Wideband-01B (J1-01B) satellite

Author

Zhang, Zhaoyi, Guo, Bofeng, Nan, Yang, Du, Hao, Li, Fenghui, Zhai, Jingsheng, Zhang, Zhaoyi, Guo, Bofeng, Nan, Yang, Du, Hao, Li, Fenghui, and Zhai, Jingsheng

Abstract

As a novel remote-sensing method, the Global Navigation Satellite System-Reflectometry (GNSS-R) can utilize a large number of reflected GNSS opportunity signals for sea ice observation. The first analysis of spaceborne GNSS-R data from the Chinese Jilin-1 Wideband-01B (J1-01B) Satellite Mission is carried out. 72 days of delay-Doppler maps (DDMs) obtained from global positioning system-reflectometry (GPS-R) and BeiDou navigation satellite system-reflectometry (BDS-R) is utilized for sea ice detection. In addition to the traditional pixel number (pn) and power summation (PS) observables, a novel DDM observable, named “trailing edge diffusion (TED),” was proposed to quantify the significant differences observed in DDMs obtained from seawater and sea ice. By establishing corresponding DDM observable thresholds, sea ice can be distinguished from seawater. Compared to the ocean and sea ice satellite application facility (OSI SAF) global sea ice concentration (SIC) product, the agreement of GPS-R and BDS-R is more than 96% and 98%, respectively, in the Antarctic. Results show high accuracy in the sea ice detection field, which demonstrates the feasibility of J1-01B DDMs and the superior performance of GPS-R and BDS-R in detecting sea ice. This study provides a novel data source for the cost-effective, data-rich, and stable GNSS-R method in sea ice detection.

Published
2024

19. Enhancing Underwater Image via Color-Cast Correction and Luminance Fusion

Author

Hu, Haofeng, Xu, Shuping, Zhao, Yazhuo, Chen, Hongyi, Yang, Shiyao, Liu, Hedong, Zhai, Jingsheng, and Li, Xiaobo

Abstract

Underwater images always suffer from color distortion, contrast decrease, and detail blur due to the selective absorption and scattering of water, which significantly limits their applications. In this article, we introduce an effective underwater image enhancement method to improve image quality, i.e., correcting color distortion, enhancing contrast, and enriching details. Specifically, we first use a color-cast factor to classify underwater images into no-color-cast and color-cast images; the latter is divided into three color shifts to help apply adaptive color correction approaches for images with different color shifts. Based on the color-corrected results, we enhance their luminance information and enrich details via a well-designed histogram equalization algorithm and a multiscale detail superposition algorithm. We finally combine fused luminance information and corrected color information to output the desired images with improved quality. Experiments on four representative underwater benchmarks validate our method's robustness to different categories of underwater images, as well as its superiority compared with state-of-the-art methods. The proposed underwater imaging solution holds significant potential for practical applications in ocean engineering, e.g., ocean vision, underwater tracking, target identification, and location.

Published
2024
Full Text
View/download PDF

38. 基于无监督学习的偏振图像去噪方法

Author

胡浩丰 Hu Haofeng, 金慧烽 Jin Huifeng, 李校博 Li Xiaobo, 翟京生 Zhai Jingsheng, and 刘铁根 Liu Tiegen

Subjects
Atomic and Molecular Physics, and Optics, Electronic, Optical and Magnetic Materials
Published
2023

47. A Semi-Analytical Method for the 3D Elastic Structural-Acoustic Radiation in Shallow Water.

Author

Qian, Zhiwen, Shang, Dejiang, Xu, Xinyang, He, Yuanan, and Zhai, Jingsheng

Subjects
WATER depth, RADIATION, ACOUSTIC field, ELASTIC plates & shells, ACOUSTIC wave propagation, FINITE element method
Abstract

A hybrid computational method is proposed for three-dimensional structural-acoustic radiation in shallow water, which combines the wave superposition method (WSM) with the adjustable Green function and the multi-physical-coupled finite element method (FEM). A low-frequency acoustic model in the near field is established by the FEM to discretize the continuous and arbitrary-shaped radiator, and the normal vibration velocity on the wetted surface is used as the input in the WSM to calculate the strengths of simple sources enclosed within the radiator. Finally, the sound propagation in the far field of the elastic structure is efficiently computed by superimposing the sound field of each simple source. The numerical examples of the pulsating sphere and elastic spherical shell in the shallow water are developed to verify the high accuracy and efficiency of the presented method, respectively. Moreover, the regularization algorithms are employed as the complementary approaches to improve the stability of this method in engineering practices, it is found that the algorithms can significantly hold the calculation accuracy in case the velocity data contain noise, and the least-squares QR is an efficient and accurate method with a few iteration numbers and better stability effect as compared with other algorithms. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results