A5005714773- Remote Sensing in Agriculture
- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Remote Sensing and LiDAR Applications
- Infrared Target Detection Methodologies
- Species Distribution and Climate Change
- Advanced Image and Video Retrieval Techniques
- Atmospheric aerosols and clouds
Huzhou University
2021-2024
University of Electronic Science and Technology of China
2021-2024
Addis Ababa Science and Technology University
2021-2022
The type of algorithm employed to classify remote sensing imageries plays a great role in affecting the accuracy. In recent decades, machine learning (ML) has received attention due its robustness image classification. this regard, random forest (RF) and support vector (SVM) are two most widely used ML algorithms generate land cover (LC) maps from satellite imageries. Although several comparisons have been conducted between these algorithms, findings contradicting. Moreover, were made on...
Cloud is a serious problem that affects the quality of remote-sensing (RS) images Existing cloud removal techniques suffer from notable limitations, such as being specific to certain data types, conditions, and spatial extents, well requiring auxiliary data, which hampers their generalizability flexibility. To address issue, we propose maximum-value compositing approach by generating masks. We acquired 432 daily MOD09GA L2 MODIS imageries covering vast region with persistent cover various...
Regional or continental-scale land cover mapping requires various amounts of months multi-temporal satellite data to pick phenological variation in vegetation, enhancing differentiability among surface types and improving accuracy. However, little has been addressed about the number months/multi-temporal images needed obtain best result impact using different these on accuracy individual classes. This work aimed analyze effects by utilizing time series FengYun-3C (FY-3C) within one year for...
This paper compared multi-temporal composite products of FY-3C VIRR and MODIS for regional land cover (LC) mapping a part Africa. LC classification was conducted using random forest algorithm in the Scikit-Learn library Python after model trained reference data that were collected by combining three techniques employed simultaneously i.e. Landsat 8 image interpretation, referring exiting maps, crosschecking on Google Earth pro/ maps. Based overall accuracy (OA) kappa value (k) two...
This work analyzes the impacts of training sample size on performance supervised classification methods when coarse resolution imageries are employed for regional land cover mapping. We utilized FegnYun-3C composite with 1km spatial and random forest (RF) support vector machine (SVM) algorithms that were trained tested five sets reference datasets: 66/34, 69/31, 73/27, 76/24 79/21.The results show two increases increasing examples until a certain point, achieves maximum accuracy (0.86 RF...