A5082859009- Automated Road and Building Extraction
- Remote-Sensing Image Classification
- Remote Sensing and LiDAR Applications
- Remote Sensing in Agriculture
- Smart Agriculture and AI
- Remote Sensing and Land Use
- Spectroscopy and Chemometric Analyses
- Machine Learning and Data Classification
- Data Management and Algorithms
- Image Processing and 3D Reconstruction
- Imbalanced Data Classification Techniques
- Landslides and related hazards
- Image and Object Detection Techniques
- Groundwater and Watershed Analysis
- Infrastructure Maintenance and Monitoring
- Industrial Vision Systems and Defect Detection
- Greenhouse Technology and Climate Control
- Water Systems and Optimization
- Advanced Neural Network Applications
- Flood Risk Assessment and Management
Zhengzhou University
2019-2025
Crop yield prediction has played a vital role in maintaining food security and been extensively investigated recent decades. Most research focused on excavating fixed spectral information from remote sensing images. However, the growth of crops is highly complex trait determined by diverse features. To maximally explore these heterogeneous features, we aim to simultaneously exploit spatial, spectral, temporal multi-spectral multi-temporal remotely sensed imagery. Therefore, this paper,...
Accurate and timely prediction of crop yield based on remote sensing data is important for food security. However, growth a complex process, which makes it quite difficult to achieve better performance. To address this problem, novel 3-D convolutional neural multikernel network proposed capture hierarchical features predicting yield. First, full constructed maximally explore deep spatial–spectral from multispectral images. Then, learning (MKL) approach fusion intraimage intersample spatial...
The precise building extraction from high-resolution remote sensing images holds significant application for urban planning, resource management, and environmental conservation. In recent years, deep neural networks (DNNs) have garnered substantial attention their adeptness in learning extracting features, becoming integral to methodologies yielding noteworthy performance outcomes. Nonetheless, prevailing DNN-based models often overlook spatial information during the feature phase....
Accurate crop classification using remote sensing imagery with limited labeled data remains a challenging yet highly valuable task in practical applications. Recently, self-supervised contrastive learning has shown considerable potential generating discriminative and generalized features from unlabeled images. Nevertheless, due to the inherent complexity of planting structures growth patterns, existing methods struggle fully capture distinct spatial spectral characteristics various crops. To...
Classification tasks on land cover (LC) mapping are challenging due to the complex and heterogeneous characteristics of remote sensing images(RSIs). Current LC classifications mainly based deep convolutional neural networks (DCNNs), previous works have been proven that spatial context can offer essential cues for performance improvement. However, they still some drawbacks limit capture ability: ambiguity global lack efficient combination strategy. To address these issues, we develop a...
Automatic land cover classification from high-resolution remote sensing (RS) images remains challenging due to the complex composition of classes. Given potential a graph simulate latent class composition, latest development convolutional network (GCN) has received increasing attention. However, most existing methods only use single perspective structure, which largely limits their ability capture complementary features that would better represent underlying data structure images. Therefore,...
The application of deep neural networks (DNNs) for road extraction from remote sensing images has gained broad interest because the competence concerning complex nonlinear relations; however, presence noisy labels in training data sets adversely affects performance DNNs. existing methods improving robustness DNNs focus on modeling noise distribution. However, these approaches are not satisfactory inaccurate high-level image features obtained by To address this issue, we develop a...
• Multiple crowdsourced data are used to reduce label noise in training samples. We propose multi-map integration model (MMIM) for road extraction.. The robustness of Deep Convolutional Neural Networks can be improved by MMIM. Best extraction accuracy achieved on a large-area covering 1059 km 2 . Road from high-resolution remote sensing images (HRSIs) is essential applications various areas. Although deep convolutional neural networks (DCNNs) have exhibited remarkable success extraction, the...
Qinghai-Tibet Plateau lakes are important carriers of water resources in the 'Asian's Water Tower', and it is great significance to grasp spatial distribution plateau for climate, ecological environment, regional cycle. However, differences spatial-spectral characteristics various types lakes, complex background information both influence extraction effect lakes. Therefore, a challenge completely effectively extract In this study, we proposed multiscale contextual aggregation network, termed...
In recent years, convolutional neural networks (CNNs) have made great achievements in object extraction from very high-resolution (VHR) images. However, most existing approaches require large quantities of clean and accurate training data to achieve impressive classification results. The presence inaccurate labels datasets is known deteriorate the performance CNNs. this paper, we introduce a novel efficient method for improving robustness when CNN on dataset with relatively noisy labels....
Recently, crowdsourced geographic data have provided a cost-effective approach to learn deep convolutional neural networks (DCNNs) for road extraction from remote-sensing images. However, datasets often suffer label noise and include error labels that can affect the performance of DCNN-based methods. Thus, we propose novel sequence learning (SDL) framework introduces probability correct robustly DCNNs extraction. The is obtained by front-end developed DCNNs, which provide valuable true...
Abstract Road extraction from high‐resolution remote sensing images (HRSIs) has great importance in various practical applications. However, most existing road methods have considerable limitation capturing long‐range shape feature of road, and thus, they are ineffective extracting region under complex scenes. To address this issue, a novel model called context‐aware neural network (LR‐RoadNet) is proposed. LR‐RoadNet takes advantage strip pooling to capture context horizontal vertical...