Annual crop inventory information is important for many agriculture applications and government statistics. The synergistic use of multi-temporal polarimetric synthetic aperture radar (SAR) available multispectral remote sensing data can reduce the temporal gaps provide spectral crops, which effective classification in areas with frequent cloud interference. main objectives this study are to develop a deep learning model map agricultural using full SAR multi-spectral data, evaluate influence...

Inspired by the tremendous success of deep learning (DL) and increased availability remote sensing data, DL-based image semantic segmentation has attracted growing interest in community. The ideal scenario DL application requires a vast number annotation data with same feature distribution as area interest. However, obtaining such enormous training sets that suit target is highly time-consuming costly. Consistency-regularization-based semi-supervised (SSL) methods have gained popularity...

Crop classification is indispensable for agricultural monitoring and food security, but early-season mapping has remained challenging. Synthetic aperture radar (SAR), such as RADARSAT Constellation Mission (RCM) Sentinel-1, can meet higher requirements on the reliability of satellite data acquisition with all-weather all-day imaging capability to supply dense observations in early crop season. This study applied local window attention transformer (LWAT) time-series SAR data, including RCM...

Compared with a monoculture planting mode, the practice of crop rotations improves fertilizer efficiency and increases yield. Large-scale rotation monitoring relies on results classification using remote sensing technology. However, limited accuracy cannot satisfy accurate identification patterns. In this paper, mapping scheme combining random forest (RF) algorithm new statistical features extracted from time-series ground range direction (GRD) Sentinel-1 images. First, synthetic aperture...

In recent years, global forest fires have occurred more frequently, seriously destroying the structural functions of ecosystem. Mapping burn severity after is great significance for quantifying fire’s effects on landscapes and establishing restoration measures. Generally, intensive field surveys across burned areas are required effective application traditional methods. Unfortunately, this requirement could not be satisfied in most cases, since work demands a lot personnel funding. For...

In the aftermath of a natural hazard, rapid and accurate building damage assessment from remote sensing imagery is crucial for disaster response rescue operations. Although recent deep learning-based studies have made considerable improvements in assessing damage, most state-of-the-art works focus on pixel-based, multi-stage approaches, which are more complicated suffer partial recognition issues at building-instance level. meantime, it usually time-consuming to acquire sufficient labeled...

Timely and accurate mapping of floodwater in urban areas from aerial imagery is critical to support emergency response rescue work. However, massive shadows cast by buildings trees over dense built-up can cause a significant underestimation flood outcomes, few studies for monitoring explore this current state-of-the-art approaches. Meanwhile, recent deep learning (DL) algorithms have reported superior performance conventional machine methods. Nevertheless, acquiring large amount training...

Objective: To investigate the blood flow change status in early stage tumor-related areas of hepatocellular carcinoma and its clinical significance after radiofrequency ablation using multi-slice spiral CT whole-liver perfusion imaging technology. Methods: 21 cases primary liver cancer that underwent CT-guided were included. scans divided into four groups according to time points (before surgery, immediately surgery 1 3 month surgery), then parameters corresponding tumor measured....

Classical machine learning models, such as linear models and tree-based are widely used in industry. These sensitive to data distribution, thus feature preprocessing, which transforms features from one distribution another, is a crucial step ensure good model quality. Manually constructing preprocessing pipeline challenging because scientists need make difficult decisions about preprocessors select order compose them. In this paper, we study how automate (Auto-FP) for tabular data. Due the...

There are many factors affect the stability of bank slopes, each them is associated and coupled with others. The analysis slopes can be achieved by method effect-factors analogy cluster analysis. Traditional difficult to obtain stable global optimal solution, since results sensitive initial center order sample input. Conventional ants clustering algorithm get an appropriate result, simulating ants’ intelligent behavior transportation. Consequently, in this paper a base on random...

Abstract On the basis of hourly rainfall dataset from 92 meteorological stations for 10 years (2008-2017), spatial and temporal distribution precipitation in Guangxi were analyzed. Results showed that high value centers found over different districts during two time frame, 6:00-12:00 18:00-24:00. The north mainly happened 18:00-24:00, but middle was 6:00-12:00. distributed unevenly on season scale. Seasonal differences diurnal variation distinguished. Summer events concentrated at morning,...