Efficient Transformer for Remote Sensing Image Segmentation

Article Agriculture, Multidisciplinary

State and parameter estimation of the AquaCrop model for winter wheat using sensitivity informed particle filter

Tianxiang Zhang et al.

Summary: Crop models play a crucial role in providing quantitative information on crop growth and field management. However, their prediction performance can be significantly affected by unknown parameters and noisy measurements. This study aims to develop an integrated dynamic SSPE framework for the AquaCrop model to maximize its potentials. The proposed framework outperforms traditional methods in both state and parameter estimation, as demonstrated in Monte Carlo simulations and a winter wheat experimental case study.

COMPUTERS AND ELECTRONICS IN AGRICULTURE (2021)

添加到收藏夹

Article Environmental Sciences

HRCNet: High-Resolution Context Extraction Network for Semantic Segmentation of Remote Sensing Images

Zhiyong Xu et al.

Summary: The paper proposes a HRCNet model based on HRNet to address the loss of spatial information and lack of global context information in conventional CNN-based semantic segmentation methods. The model integrates LDA and FEFP structures to fuse contextual information of different scales and BA module combined with BAloss function to achieve boundary information. Experimental results show significant improvement in boundary and segmentation performance on Potsdam and Vaihingen datasets, with overall accuracy scores increased up to 92.0% and 92.3% respectively.

REMOTE SENSING (2021)

添加到收藏夹

Article Chemistry, Multidisciplinary

Sentinel-2 Satellite Imagery for Urban Land Cover Classification by Optimized Random Forest Classifier

Tianxiang Zhang et al.

Summary: This study proposes a systematic method to automatically tune the hyperparameters of a random forest classifier by Bayesian parameter optimization, and utilizes Sentinel-2A/B satellite imagery for land cover classification in Beijing. The experimental results demonstrate that the optimized random forest classifier outperforms traditional support vector machines and random forests with default hyperparameters in terms of accuracy, precision, and recall.

APPLIED SCIENCES-BASEL (2021)

添加到收藏夹

Article Environmental Sciences

Boundary-Aware Refined Network for Automatic Building Extraction in Very High-Resolution Urban Aerial Images

Yuwei Jin et al.

Summary: BARNet is a novel network with a special boundary-aware loss that improves the segmentation performance of large-scale buildings using gated-attention refined fusion unit and denser atrous spatial pyramid pooling module. Experimental results demonstrate that BARNet outperforms several state-of-the-art approaches in both visual interpretation and quantitative evaluations.

REMOTE SENSING (2021)

添加到收藏夹

Article Environmental Sciences

Semantic Segmentation of Aerial Imagery via Split-Attention Networks with Disentangled Nonlocal and Edge Supervision

Cheng Zhang et al.

Summary: In this work, a new deep convolution neural network architecture is proposed for semantic segmentation of aerial imagery, leveraging recent research and combining various modules to improve segmentation accuracy.

REMOTE SENSING (2021)

添加到收藏夹

Article Environmental Sciences