A5100436443- Atmospheric aerosols and clouds
- Image and Signal Denoising Methods
- Meteorological Phenomena and Simulations
- Advanced Image Processing Techniques
- Medical Image Segmentation Techniques
- Atmospheric chemistry and aerosols
- Geology and Paleoclimatology Research
- Image Enhancement Techniques
- Robotic Path Planning Algorithms
- Precipitation Measurement and Analysis
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Robotics and Sensor-Based Localization
- Advanced Image Fusion Techniques
- Computational Physics and Python Applications
- Marine and environmental studies
- Sentiment Analysis and Opinion Mining
- Guidance and Control Systems
- Archaeology and ancient environmental studies
- Air Quality Monitoring and Forecasting
- Anodic Oxide Films and Nanostructures
- Grouting, Rheology, and Soil Mechanics
- Ideological and Political Education
- Geoscience and Mining Technology
- Image and Object Detection Techniques
Chengdu University of Information Technology
2018-2024
Harbin University of Science and Technology
2022-2024
Tohoku University
2023
State Grid Corporation of China (China)
2022
Nanjing University of Information Science and Technology
2019-2022
Dalian Ocean University
2022
Yangzhou University
2022
China Meteorological Administration
2020-2021
Northwestern Polytechnical University
2021
Northeast Forestry University
2020
Abstract Ice clouds are mostly composed of different ice crystal habits. It is great importance to classify habits seeing as they could greatly impact single‐scattering properties particles. The play an important role in the study cloud remote sensing and Earth's atmospheric radiation budget. However, there countless crystals with shapes clouds, task empirical classification based on naked‐eye observations unreliable, time consuming subjective, which leads results having obvious...
In the atmosphere, cloud particles have different shapes. The study of particle shapes plays an important role in understanding precipitation processes, radiative transfer, and weather modification. image resolution data quality probes affect accuracy classification To solve occlusion photosensitive edge achieve automatic, high-precision ice-crystal airborne Cloud Imaging Probe (CIP) images, this uses a traditional processing algorithm for control applies artificial intelligence algorithms...
<abstract> <p><italic>Purpose</italic>: Due to the complex distribution of liver tumors in abdomen, accuracy tumor segmentation cannot meet needs clinical assistance yet. This paper aims propose a new end-to-end network improve from CT. <italic>Method</italic>: We proposed hybrid network, leveraging residual block, context encoder (CE), and Attention-Unet, called ResCEAttUnet. The CE comprises dense atrous convolution (DAC) module multi-kernel pooling...
Image priors have been successfully introduced to solve ill-posed problems, such as image denoising. In this paper, we propose a new denoising model for magnetic resonance images (MRIs) which employs the low-rank and sparse gradient priors. First, use Gaussian mixture (GMM) guide clustering of non-local self-similar patches by learning structure external noise-free MRI help retain low rank noisy patch matrix. Second, fit heavy-tailed with hyper-Laplacian distribution reduce ringing...
The combination of imaging information from multi-modality images may be highly beneficial for radiotherapy treatment planning in terms tumor delineation. This paper proposes a discriminative learning based approach automated nasopharyngeal carcinoma (NPC) segmentation using images. Specially, an image-patch-based convolutional neural network (CNN) is designed to jointly learn similarity metric and classification paired image patch different modalities. CNN integrates two normal sub-networks...
Abstract It is not denied that real-time monitoring of radar products an important part in actual meteorological operations. But the weather often brings out abnormal echoes due to various factors, such as climate and hardware failure. So it great practical significance research value realize automatic identification anomaly products. However, traditional algorithms identify anomalies echo images are most accurate efficient. In order improve efficiency identification, a novel method...
Research Article| March 01, 2015 Mapping and measuring Lake Agassiz strandlines in North Dakota Manitoba using LiDAR DEM data: Comparing techniques, revising correlations, interpreting anomalous isostatic rebound gradients James T. Teller; Teller † 1Department of Geological Sciences, University Manitoba, Winnipeg, R3T 2N2, Canada †tellerjt@ms.umanitoba.ca. Search for other works by this author on: GSW Google Scholar Zhirong Yang 2Water Survey Canada, Environment 373 Sussex Drive, Ottawa,...
Abstract Satellite cloud images play an important role in weather analysis and forecast. High‐resolution satellite a significant the study of mesoscale systems such as typhoons. With increasing demands locating tracking techniques, resolution is no longer satisfactory. Enhancing their with superresolution (SR) methods can help identifying systems. In this paper, we propose multipath network model, called SRCloudNet, that involves joint training back‐projection local residual network....
Abstract Cloud particles have different shapes in the atmosphere. Research on cloud particle plays an important role analyzing growth of ice crystals and microphysics. To achieve accurate efficient classification algorithm crystal images, this study uses image-based morphological processing principal component analysis, to extract features images apply intelligent algorithms for Particle Imager (CPI). Currently, there are mainly two types ice-crystal methods: one is mode parameterization...
A new fuzzy level set method (FLSM) based on the global search capability of quantum particle swarm optimization (QPSO) is proposed to improve stability and precision image segmentation, reduce sensitivity initialization. The combination QPSO-FLSM algorithm iteratively optimizes initial contours using QPSO c-means clustering, then utilizes (LSM) segment images. exploits obtain a stable cluster center pre-segmentation contour closer region interest during iteration. In implementation in...
The color change of opal photonic crystal films (OPCFs) due to deformation was quantitatively evaluated using digital image correlation (DIC) analysis. OPCFs were pasted on specimens three different gauge geometries, and random patterns formed the opposite side each specimen for DIC To assess applicability OPCFs-based strain characterization analyzing steel structural components associated metallurgical analyses, smooth, width-gradient, holed prepared in this study. As increased smooth...
The development of a general purpose service robot for daily life necessitates the robot's ability to deploy myriad fundamental behaviors judiciously. Recent advancements in training Large Language Models (LLMs) can be used generate action sequences directly, given an instruction natural language with no additional domain information. However, while outputs LLMs are semantically correct, generated task plans may not accurately map acceptable actions and might encompass various linguistic...
Accurate spare part demand forecasting for the key components of ships is one factors to ensure normal ship operation. Large errors in not only bring challenges operation but also cause an inventory backlog and thus increase maintenance costs. A three-level parts combination prediction method based on historical data has been proposed, aiming solve problem insufficient existing methods. First, three types individual direct models are used predictions. Secondly, we convolutional neural...
<title>Abstract</title> Current neural network-based optic disc (OD) and cup (OC) segmentation tend to prioritize the image's local edge features, thus limiting their capacity model long-term relationships, with errors in delineating boundaries. To address this issue, we proposed a semi-supervised Dual Self-Integrated Transformer Network (DST-Net) for joint of OD OC. Firstly, construct encoder decoder self-integrated network from mutually enhanced feature learning modules Vision (ViT)...
Digital image processing is a course with strong theoretical backgrounds and also has wide applications in reality. However, traditional teaching method only focuses on delivering abstract theories, which leads to decrease students' interest the course. To resolve this problem, paper studies characteristics of basic algorithms from perspective combining theory practice. We investigate fundamental algorithms. implement some cases using Python.
Precipitation with high spatial and temporal resolution can improve the defense capability of meteorological disasters provide indispensable instruction early warning for social public services, such as agriculture, forestry, transportation. Therefore, a deep learning-based algorithm entitled precipitation retrieval from satellite observations based on Transformer (PRSOT) is proposed to fill observation gap ground rain gauges weather radars in deserts, oceans, other regions. In this...
The K work area belongs to the Chalju terrace, north of Amu Basin in structure. With development from exploration and evaluation stage gradually, requirements for reservoir prediction are getting higher. To improve accuracy prediction, paper has chosen acoustic impedance as physical property sensitive parameter carbonate based on results four-characters research predicted plane distribution gas using waveform indication inversion technique. technology seismic indicated which instructed by...
Abstract Explaining the practice of social development shows that bearing capacity is a problem must be faced by regional sustainable development. This paper uses restrictive factor method to calculate and analyze Dianchi Lake Resort. The results show Kunming National Tourism Resort currently in state overall coordination some problems are outstanding; corresponding suggestions proposed.