GF-1 Images Research Articles

Change detection of wetland restoration in China’s Sanjiang National Nature Reserve using STANet method based on GF-1 and GF-6 images

Wetlands are important ecosystems, and detecting changes in wetlands plays a very important role in the management of wetland resources. Therefore, we selected Sanjiang National Nature Reserve as the study area and proposed a method in which bi-temporal GaoFen (GF) remote sensing data was combined with a spatial–temporal attention neural network (STANet) model, coupled with the selection of spectral variables. We used the bi-temporal images from GF-1 and GF-6 and compared three band selection methods, Relief F, principal component analysis, and RGB combination, and two STANet-based change detection methods (BAM: basic spatial–temporal attention module and PAM: pyramid spatial–temporal attention module) to improve the change detection of the wetland restoration. The results show that the accuracy of wetland change detection is related to different combinations of spectral variable selection and STANet models. The combination of Relief F feature selection and the BAM model led to the highest accuracy of the wetland change detection, 71.72 % for F1-Score and 76.92 % for MIou. The vector results of the change detection statistics show that the area of the Sanjiang National Nature Reserve cultivated lands converted to wetlands is 46.05 km2. This study implies that the STANet method has the capacity of detecting complex wetland restoration. Red, Bule, NIR and Red Edge1 bands of GaoFen data can provide effective information in the detection of changes in wetlands. The results of the study can be used as a reference for rational planning of wetland restoration, seeking a balance between cultivated lands and wetlands, which is important for the sustainable development of wetland.

Multiscale spatiotemporal dynamics analysis of urban green space: Implications for green space planning in the rapid urbanizing Hefei City, China

Under the pressure of rapid urbanization, the spatiotemporal dynamics of urban green spaces (UGS) have enormous impacts on the local ecological system and environment at different scales. In this study, UGS in Hefei City, which has experienced rapid urbanization from 1995 to 2015, were extracted based on time-series Landsat-5 TM and Landsat-8 OLI images, and different types of parks were mapped based on GF-2 images combined with multi-source metadata. Dynamic patterns of green space were examined by drawing the spatial variations of green space at the city, inner city, and park scales. Results revealed that: (1) At the city level, UGS decreased with the rapid urbanization, especially farmland sharply lost in areas around existing urban areas and along the transportation corridors. (2) At the inner-city level, concentric analyses showed that UGS changes in different rings had unique trends, and the dynamic changes were the most significant in Rings 2 and 3. Comparative analysis of old and new districts showed that the common characteristics of UGS changes were the transfer-out of farmland and the transfer-in of built-up land, and the newer the region, the more intense the changes. (3) At the park level, the number and area of urban parks were increasing from the center to the periphery, and the type of park gradually changed from single to rich. Significant spatial heterogeneity was identified in the landscape patterns of urban parks based on fishnet cell analysis. It is fundamental to assign urban construction land for socio-economic development, while planning UGS at different scales; moreover, integrating different green space-related policies could protect the UGS and maintain its stability. Only in this way can UGS combine economic, social, and ecological benefits under the background of rapid urbanization.

Accurate Recognition of Building Rooftops and Assessment of Long-Term Carbon Emission Reduction from Rooftop Solar Photovoltaic Systems Fusing GF-2 and Multi-Source Data

Rooftop solar photovoltaic (PV) retrofitting can greatly reduce the emissions of greenhouse gases, thus contributing to carbon neutrality. Effective assessment of carbon emission reduction has become an urgent challenge for the government and for business enterprises. In this study, we propose a method to assess accurately the potential reduction of long-term carbon emission by installing solar PV on rooftops. This is achieved using the joint action of GF-2 satellite images, Point of Interest (POI) data, and meteorological data. Firstly, we introduce a building extraction method that extends the DeepLabv3+ by fusing the contextual information of building rooftops in GF-2 images through multi-sensory fields. Secondly, a ridgeline detection algorithm for rooftop classification is proposed, based on the Hough transform and Canny edge detection. POI semantic information is used to calculate the usable area under different subsidy policies. Finally, a multilayer perceptron (MLP) is constructed for long-term PV electricity generation series with regional meteorological data, and carbon emission reduction is estimated for three scenarios: the best, the general, and the worst. Experiments were conducted with GF-2 satellite images collected in Daxing District, Beijing, China in 2021. Final results showed that: (1) The building rooftop recognition method achieved overall accuracy of 95.56%; (2) The best, the general and the worst amount of annual carbon emission reductions in the study area were 7,705,100 tons, 6,031,400 tons, and 632,300 tons, respectively; (3) Multi-source data, such as POIs and climate factors play an indispensable role for long-term estimation of carbon emission reduction. The method and conclusions provide a feasible approach for quantitative assessment of carbon reduction and policy evaluation.

Comparison of the use of Landsat 8, Sentinel-2, and Gaofen-2 images for mapping soil pH in Dehui, northeastern China

Mapping the distribution and quantity of soil properties is important for black soil protection, management, and restoration in northeastern China. The objective of this study was to evaluate the effect of the spatial resolution on soil pH mapping using satellite images of the black soil region in northeastern China. A high spatial resolution Gaofen (GF)-2 high-definition image and multispectral images acquired by the Landsat 8 operational land imager and Sentinel-2 multi-spectral instrument were used to compare their performance in soil pH prediction. The spectral variables, including the original bands of the three satellite images and a variety of spectral indices derived from the original bands, were employed. Then, a machine learning model (quantile regression forest) was used to determine the relationships between the spectral variables and the measured soil pH, and prediction models were established to estimate the soil pH and to characterize the spatial pattern of the soil pH. The results revealed that the soil pH prediction model based on the GF-2 image had a slightly higher prediction accuracy than the models constructed using the Landsat 8 and Sentinel-2 images. The prediction models for Landsat 8, Sentinel-2, and GF-2 had root mean square errors of 0.34, 0.39, and 0.31, respectively. The use of remote sensing images with a high spatial resolution may not substantially increase the prediction accuracy of soil pH mapping compared with the results derived from medium-resolution images.

DEEP LEARNING APPROACH FOR FLOOD DETECTION USING SAR IMAGE: A CASE STUDY IN XINXIANG

Abstract. With the gradual warming of the global climate, frequent floods have caused huge losses to human life and property. Flood mapping by SAR image has been an important topic, and it is increasingly important to use deep learning method to extract flood information. In order to achieve automatic flood extent extraction, this paper proposes an attention mechanism-based water body extraction network with GF3 images, and successfully use it for flood detection in Xinxiang, Henan, China. In this paper, the proposed network incorporates the channel attention mechanism and position attention mechanism based on U-Net, to improve the efficiency and accuracy of water extraction, ignore the unimportant information by learning the weight and make the model focus toward the important information related to the water body. The OA of our method can reach 0.959 and Recall can reach 0.942 by verifying four sets of test data. Experiments show fast flood mapping in Xinxiang can be achieved by our method.

Automatic Extraction of Mountain River Surface and Width Based on Multisource High-Resolution Satellite Images

The extraction of high-resolution geomorphic information from remote sensing images is a key technology for supporting mountain river research. Extracting small rivers (width < 90 m) from complex backgrounds based on satellite images remains a challenging issue. In this research, we propose an improved random forest (RF) algorithm, RF-ANN (artificial neural network), by using neural networks and thermal infrared data for the extraction of river surfaces. We also develop an automated river width extraction (ARWE) method based on the central axis transformation algorithm and centerline automatic correction algorithm for the automatic extraction of the river widths across the whole basin. We chose the Huangfuchuan River Basin on the Loess Plateau, China, as a case study area. Chinese GF-1 and ZY-3 satellite images were implemented as the primary data source. We extracted the bankfull river surface and river widths of the Huangfuchuan River by using these two improved methods. The results show that the RF-ANN method has a total river surface extraction accuracy of 94.7%, and the extracted river surfaces cover more than 85% of the order 3 DEM river network. By implementing high-resolution DEM and thermal infrared data, RF-ANN effectively eliminates the disturbance of shadows of mountains and other features, which ensures the high accuracy of the extracted widths. It was verified that the maximum and minimum river widths that can be extracted in the Huangfuchuan River Basin are 297.4 m and 6.1 m, respectively. The overall error of river width extraction is 0.97 m, which is less than half of the pixel length of remote sensing images. The R2 and root mean square error (RMSE) of the estimated river width values are 0.99 and 1.49, respectively. For tiny rivers with widths narrower than 10 m, the error of river width extraction is 10.9%. The error of thin rivers whose widths range from 10 to 30 m is 4.9%. For small rivers ranging from 30 to 90 and rivers wider than 90 m, the error is 1.1% and 0.6%, respectively. The new approach provides an effective method for extracting the surface and width of mountain rivers in topographically complex regions by using high-resolution satellite images, which may provide a database for estimating river carbon emissions and related research in fluvial morphology and water resource management.

Comparing the Sensitivity of Pixel-Based and Sub-Watershed-Based Analytic Hierarchy Process to Weighting Criteria for Flood Hazard Estimation

In flood hazard estimation via the analytic hierarchy process (AHP), using the pixel as the basic unit might lead to accuracy relying on the optimal weighting criteria. To this end, considering the sub-watershed as the basic unit is new. In this study, taking the Chaohu Basin in Anhui Province, China, as a study case, the accuracy of the sensitivity of the pixel-based and sub-watershed-based AHP models influenced by weighting criteria was compared. There were 48 judgment ma- trixes defined, following the same order of importance of the involved indicators. Validation ground truthing is constructed by the extracted flooded regions from GF-3 images. As weighting criteria changed, the results indicated that the pixel-based AHP fluctuated significantly, while the correct ratio and fit ratio derived by the sub-watershed-based AHP could improve by >35% and >5%, respectively, over the pixel-based-AHP. It indicated that the sub-watershed-based AHP has an advantage in relying less on in situ weighting criteria than the pixel-based AHP.

Estimation of Winter Wheat Residue Coverage Based on GF-1 Imagery and Machine Learning Algorithm

Crop residue is an important component of farmland ecosystems, which is of great significance for increasing soil organic carbon, mitigating wind erosion and water erosion and conserving soil and water. Crop residue coverage (CRC) is an important parameter to characterize the number and distribution of crop residues, and also a key indicator of conservation tillage. In this study, the CRC of wheat was taken as the research object. Based on the high-resolution GF-1 satellite remote sensing imagery from China, decision tree (DT), gradient boosting decision tree (GBDT), random forest (RF), least absolute shrinkage and selection operator (LASSO), extreme gradient boosting regression (XGBR) and other machine learning algorithms were used to carry out the estimation of wheat CRC by remote sensing. In addition, the comparisons with sentinel-2 imagery data were also utilized to assess the potential of GF satellite data for CRC estimates. The results show the following: (1) Among the spectral indexes using shortwave infrared characteristic bands from sentinel-2 imagery, the dead fuel index (DFI) was the best for estimating wheat CRC, with an R2 of 0.54 and an RMSE of 10.26%. The ratio vegetation index (RVI) extracted from visible and near-infrared characteristic bands from GF-1 data performed the best, with an R2 of 0.46 and an RMSE of 11.39%. The spectral index extracted from GF-1 and sentinel-2 images had a significant response relationship with wheat residue coverage. (2) When only the characteristic bands from the visible and near-infrared spectral ranges were applied, the effects of the spatial resolution differences of different images on wheat CRC had to be taken into account. The estimations of wheat CRC with the high-resolution GF-1 data were significantly better than those with the Sentinel-2 data, and among multiple machine learning algorithms adopted to estimate wheat CRC, LASSO had the most stable capability, with an R2 of 0.46 and an RMSE of 11.4%. This indicates that GF-1 high-resolution satellite imagery without shortwave infrared bands has a good potential in applications of monitoring crop residue coverage for wheat, and the relevant technology and method can also provide a useful reference for CRC estimates of other crops.

Monitoring Potential Geological Hazards with Different InSAR Algorithms: The Case of Western Sichuan

In recent years, the number of geological disasters in Sichuan Province has significantly increased due to the influence of earthquakes and extreme climate, as well as the disturbance to the geological environment by human activities. Thus, geological disaster monitoring is particularly important, which can provide some scientific basis for disaster prevention and reduction. In this paper, the interferometric synthetic aperture radar (InSAR) technology was introduced to monitor potential geological hazards, taking parts of Dujiangyan City, Wenchuan County, and Mao County in Sichuan Province, China as examples. Firstly, the data such as Sentinel-1A Terrain Observation with Progressive Scans (TOPS) Synthetic Aperture Radar (SAR) images and Precision Orbit Determination (POD) precise orbit ephemerides from 2018 to 2020, high-resolution optical satellite images and Digital Elevation Model (DEM) were collected. Secondly, the Differential InSAR (D-InSAR), Persistent Scatterer InSAR (PS-InSAR), Small Baseline Subset InSAR (SBAS-InSAR), Offset-Tracking, and Distributed Scatterer InSAR (DS-InSAR) algorithms were used to invert the surface deformation of the study area. Thirdly, according to the deformation results obtained by experiments, we used GF-1 and GF-2 optical images as a reference and combine the experimental results of InSAR algorithms to delineate the areas affected by geological disasters. A total of 49 geological disaster areas were obtained, mainly including landslides, collapses, and debris flow. Through field verification, the overall accuracy rate of InSAR deformation monitoring achieved 69.23%, and the accuracy rate of new potential hazards monitoring reached 63.64%. Among all InSAR methods, the DS-InSAR method outperformed and monitored the geological disaster areas well. Finally, the study area was divided into three elevation intervals and the applicability of different InSAR algorithms in different elevation intervals was discussed.

Optimal spatial resolution of remote-sensing imagery for monitoring cantaloupe greenhouses

Plastic greenhouses are vital agricultural facilities to protect cash crops from disease and insects, especially in the Hainan region of China, which has high temperature and high humidity. Remote-sensing technology is an efficient means to quickly determine the spatial distribution of plastic greenhouses on the regional scale. With the rapid development of remote-sensing technology, and especially the increasing types of high-spatial-resolution remote-sensing imagery, many studies have obtained good results by using remote-sensing technology to monitor plastic greenhouses. However, the best spatial resolution of images for monitoring plastic greenhouses has yet to be studied. To address this issue, we use cantaloupe greenhouses as the research object and GF-2 images with 1m spatial resolution as data source. We then use the re-sampling method to generate images from these data with spatial resolutions of 0.5, 2, 3, and 5 m. The details of the spatial distribution (texture features and shape features) and the spectral features of the plastic greenhouses were then extracted from images of varying spatial resolution, and a remote-sensing monitoring method for cantaloupe greenhouses was constructed based on the object-oriented random forest algorithm, which combines spectral, texture and shape features, and the monitoring results are compared. The results show that the use of 2 m spatial resolution provides the highest monitoring accuracy of cantaloupe greenhouses (overall accuracy = 94.85% and KIA = 0.92). This study thus provides a theoretical basis for remote-sensing monitoring of greenhouse cantaloupes that satisfies the current demands of production accuracy.

Stepwise Fusion of Hyperspectral, Multispectral and Panchromatic Images with Spectral Grouping Strategy: A Comparative Study Using GF5 and GF1 Images

Since hyperspectral satellite images (HSIs) usually hold low spatial resolution, improving the spatial resolution of hyperspectral imaging (HSI) is an effective solution to explore its potential for remote sensing applications, such as land cover mapping over urban and coastal areas. The fusion of HSIs with high spatial resolution multispectral images (MSIs) and panchromatic (PAN) images could be a solution. To address the challenging work of fusing HSIs, MSIs and PAN images, a novel easy-to-implement stepwise fusion approach was proposed in this study. The fusion of HSIs and MSIs was decomposed into a set of simple image fusion tasks through spectral grouping strategy. HSI, MSI and PAN images were fused step by step using existing image fusion algorithms. According to different fusion order, two strategies ((HSI+MSI)+PAN and HSI+(MSI+PAN)) were proposed. Using simulated and real Gaofen-5 (GF-5) HSI, MSI and PAN images from the Gaofen-1 (GF-1) PMS sensor as experimental data, we compared the proposed stepwise fusion strategies with the traditional fusion strategy (HSI+PAN), and compared the performances of six fusion algorithms under three fusion strategies. We comprehensively evaluated the fused results through three aspects: spectral fidelity, spatial fidelity and computation efficiency evaluation. The results showed that (1) the spectral fidelity of the fused images obtained by stepwise fusion strategies was better than that of the traditional strategy; (2) the proposed stepwise strategies performed better or comparable spatial fidelity than traditional strategy; (3) the stepwise strategy did not significantly increase the time complexity compared to the traditional strategy; and (4) we also provide suggestions for selecting image fusion algorithms using the proposed strategy. The study provided us with a reference for the selection of fusion strategies and algorithms in different application scenarios, and also provided an easy-to-implement solution and useful references for fusing HSI, MSI and PAN images.

Analysis on ecological status and spatial-temporal variation of Tamarix chinensis forest based on spectral characteristics and remote sensing vegetation indices.

The reserve of Tamarix forest, located in Changyi, China, is the only national marine special reserve taking Tamarix as the main object of protection. Compared with conventional monitoring technology, remote sensing technology can more comprehensively reflect the ecological environment status and spatial-temporal variation of monitoring objects. Based on spectral characteristics and remote sensing vegetation indices, the ecological status and spatial-temporal variation of Tamarix chinensis forest in the reserve deserve further exploration. Therefore, spectral characteristic, typical vegetation indices, comprehensive health index, VFC, and REP were analyzed based on Sentinel-2A images. Spatial-temporal variation analysis during 2014 to 2018 was analyzed based on GF-1 images. The research result indicated that ecological quality of protection zone showed an overall growth trend with the help of artificial ecological restoration, and it is possible to continuously implement ecological recovery towards the protection zone.

Extracting Lakes and Reservoirs From GF-1 Satellite Imagery Over China Using Improved U-Net

Lakes and reservoirs (LaR) are important parts of water resources and their rapid and accurate monitoring is an essential guarantee for maintaining ecological health and social development. The existing waterbody extraction methods are mostly targeted at local water bodies, with little attention on the national scale. In this letter, an improved U-Net method is proposed for LaR extraction from GF-1 satellite imagery. First, 21 scenes of GF-1 images are evenly selected across China, and the training set and validation set are produced by image processing, cropping, and augmentation. Second, a deep learning network is constructed by modifying the U-Net, deepening the network and introducing multiple skip connections, which is suitable for extracting LaR China-wide. Experiments on the GF-1 imagery demonstrate that the superiority of the improved U-Net when compared with other deep learning methods (U-Net, UNet++, FastFCN, DeepLabv3+) and traditional methods [the normalized difference water index (NDWI), maximum likelihood method (MLM)]. In addition, 20 LaR are selected for further evaluation of the model, and all of them achieve good extraction results, showing excellent generalization of the model.

Brightness Compensation for GF-3 SAR Images Based on Cycle-Consistent Adversarial Networks

GF-3 is the first C-Band full-polarimetric synthetic aperture radar (SAR) satellite with a space resolution up to 1m in China. The uneven brightness of SAR images is a problem when using GF-3 images, which makes it difficult to use and produce SAR images. In this paper, a brightness compensation method is proposed for GF-3 SAR images with unbalanced brightness in some areas based on a deep learning model named Cycle-Consistent Adversarial Networks (CycleGAN). The proposed method makes the image brightness relatively consistent, and it is compared with the MASK dodging algorithm, Wallis dodging algorithm and histogram equalization in terms of the profiles, brightness mean, standard deviation, and average gradient. Results of brightness compensation show that, the proposed method makes the inner brightness differences smaller, and the image quality is obviously improved, which provides even brightness image for subsequent applications, and has great practical significance.

An Adaptive-Parameter Pixel Unmixing Method for Mapping Evergreen Forest Fractions Based on Time-Series NDVI: A Case Study of Southern China

Spectral unmixing remains the most popular method for estimating the composition of mixed pixels. However, the spectral-based unmixing method cannot easily distinguish vegetation with similar spectral characteristics (e.g., different forest tree species). Furthermore, in large areas with significant heterogeneity, extracting a large number of pure endmember samples is challenging. Here, we implement a fractional evergreen forest cover-self-adaptive parameter (FEVC-SAP) approach to measure FEVC at the regional scale from continuous intra-year time-series normalized difference vegetation index (NDVI) values derived from moderate resolution imaging spectroradiometer (MODIS) imagery acquired over southern China, an area with a complex mixture of temperate, subtropical, and tropical climates containing evergreen and deciduous forests. Considering the cover of evergreen forest as a fraction of total forest (evergreen forest plus non-evergreen forest), the dimidiate pixel model combined with an index of evergreen forest phenological characteristics (NDVIann-min: intra-annual minimum NDVI value) was used to distinguish between evergreen and non-evergreen forests within a pixel. Due to spatial heterogeneity, the optimal model parameters differ among regions. By dividing the study area into grids, our method converts image spectral information into gray level information and uses the Otsu threshold segmentation method to simulate the appropriate parameters for each grid for adaptive acquisition of FEVC parameters. Mapping accuracy was assessed at the pixel and sub-pixel scales. At the pixel scale, a confusion matrix was constructed with higher overall accuracy (87.5%) of evergreen forest classification than existing land cover products, including GLC 30 and MOD12. At the sub-pixel scale, a strong linear correlation was found between the cover fraction predicted by our method and the reference cover fraction obtained from GF-1 images (R2 = 0.86). Compared to other methods, the FEVC-SAP had a lower estimation deviation (root mean square error = 8.6%). Moreover, the proposed method had greater estimation accuracy in densely than sparsely forested areas. Our results highlight the utility of the adaptive-parameter linear unmixing model for quantitative evaluation of the coverage of evergreen forest and other vegetation types at large scales.

Forest classification using synthetic GF-1/WFV time series and phenological parameters

High-accuracy forest type mapping is crucial for forest resource inventory to support forest management, conservation biology, and ecological restoration. The Chinese GF-1 satellite can provide high-spatial resolution and low-cost data for forestry monitoring. However, it is difficult to obtain time-series GF-1 images with high spatial resolution and high temporal frequency because of technical constraints and cloud contamination, which affects the accuracy of forest type extraction. We proposed a method for mapping forest type using Chinese GF-1/widefield view (WFV) multi-spectral data, the fused GF-1/WFV normalized difference vegetation index (NDVI) time series, and the phenological parameters. We used the enhanced spatial and temporal adaptive reflectance fusion model algorithm to fuse Chinese GF-1/WFV and Moderate Resolution Imaging Spectroradiometer NDVI to generate high-spatiotemporal resolution NDVI time-series image data and extract phenological parameters in the northeastern part of Hubei Province, China. We then designed four different combinations of GF-1/WFV spectral features, the fused NDVI time-series features, and the phenological parameters. Finally, a random forest algorithm is used to classify forest types in the study area. The results showed that the proposed method can obtain satisfactory results. The overall accuracy based on the combination of GF-1/WFV spectral features, high-resolution time-series NDVI, and phenological features was 89.68%. Compared with the classification using only GF-1/WFV spectral data, the overall accuracy was improved by 9.60% points. These results demonstrate that the proposed method can support forest type survey and mapping at the regional or global scale.

Multi-resolution satellite images bathymetry inversion of Bangda Co in the western Tibetan Plateau

ABSTRACT Water depth is important information for lake research. However, it is difficult to obtain the depth distribution of a whole lake. The method of remote-sensing inversion by combining bathymetric data has been proved to be feasible. In order to improve the inversion accuracy, it is very crucial to research the optimal inversion model and the most suitable depth for modelling, and the most appropriate satellite images for the study area. In this study, we used the three kinds of multispectral image data (Sentinel-2, GF-1, and Landsat-8) to establish models by combining bathymetric data in Bangda Co, which is located in the northwestern Tibetan Plateau. We verified the two empirical models (Stumpf model and Lyzenga model) with the three bands of blue, green, and red in 0–30 m water depth. We created a multi-factor combination model (S-L model) to compare with a BP neural network model using three images in five different water depth ranges (0–10, 0–15, 0–20, 0–25, and 0–30 m). The results indicated that the S-L model with mean absolute error (MAE) of 3.09–4.70 m, which the accuracy was slightly better than that of the two empirical models (3.35–5.03 and 3.23–4.94 m). The MAE of the BP model over the five depth ranges was 0.71–2.41 m for Landsat-8, 1.02–3.53 m for GF-1, and 1.15–3.68 m for Sentinel-2, which was higher than the S-L model. For the five ranges of three images by using the BP model, both the accuracy of Sentinel-2 and GF-1 in the range of 0–15 m was higher, but Landsat-8 was the highest at 0–20 m among five ranges; the mean relative errors (MRE) were 19.5%, 18.7%, 13.1%, and the root-mean-square error (RMSE) were 1.82, 1.72, and 1.86 m, respectively. The best applicability of the three images was Landsat-8, followed by GF-1 and Sentinel-2 in this study. The results suggested that the freely available Sentinel-2, GF-1 and Landsat-8 images could be used to estimate water storage for a shallow lake by combining bathymetric data on the Tibetan Plateau.

A Region-Based Deep Reinforcement Learning Classification Method for GF-3 PolSAR Imagery Classification

The C-band GF-3 satellite is the first fully polarimetric synthetic aperture (PolSAR) designed by China, which has multi-polarization image modes and been used in various applications. Land cover classification is an effective approach in PolSAR image interpretation and further application. However, conventional classification methods are mostly pixel-based and are easily affected by inherent speckle noise. In addition, the feature selection of PolSAR image and the amount of training samples are also critical to the performance of classifiers. To solve these problems, in this paper, we propose a region-based PolSAR image classification method, which uses reinforcement learning method altogether with statistical region merging algorithm to improve the classification performance. The contributions of our method are mainly reflected in three aspects: First, the T3 matrix is considered as the only feature set in our method, including image segmentation and classification. Second, the region is produced via statistical region merging algorithm. Finally, a deep reinforcement learning model is used to obtain PolSAR image classification result. To evaluate the performance of the proposed method, two real GF-3 images are performed in the experiments, and the experimental results illustrate that the proposed method outperforms the conventional methods (support vector machine, random forest, and convolution neural network) in terms of accuracy and achieves the state-of-art results.

How to Measure the Urban Park Cooling Island? A Perspective of Absolute and Relative Indicators Using Remote Sensing and Buffer Analysis

Urban parks have been proven to cool the surrounding environment, and can thus mitigate the urban heat island to an extent by forming a park cooling island. However, a comprehensive understanding of the mechanism of park cooling islands is still required. Therefore, we studied 32 urban parks in Jinan, China and proposed absolute and relative indicators to depict the detailed features of the park cooling island. High-spatial-resolution GF-2 images were used to obtain the land cover of parks, and Landsat 8 TIR images were used to examine the thermal environment by applying buffer analysis. Linear statistical models were developed to explore the relationships between park characteristics and the park cooling island. The results showed that the average land surface temperature (LST) of urban parks was approximately 3.6 °C lower than that of the study area, with the largest temperature difference of 7.84 °C occurring during summer daytime, while the average park cooling area was approximately 120.68 ha. The park cooling island could be classified into four categories—regular, declined, increased, and others—based on the changing features of the surrounding LSTs. Park area (PA), park perimeter (PP), water area proportion (WAP), and park shape index (PSI) were significantly negatively correlated with the park LST. We also found that WAP, PP, and greenness (characterized by the normalized difference vegetation index (NDVI)) were three important factors that determined the park cooling island. However, the relationship between PA and the park cooling island was complex, as the results indicated that only parks larger than a threshold size (20 ha in our study) would provide a larger cooling effect with the increase in park size. In this case, increasing the NDVI of the parks by planting more vegetation would be a more sustainable and effective solution to form a stronger park cooling island.

A Comparative Research on Three Identification Methods of Scenic Resources of Tourism Based on GF-2 Image: A Case Study of Yesanpo National Park

Based on the scientific identification and evaluation of scenic resources, the formulation of planning and management is of great significance to the sustainable development of tourism in national parks. This paper uses GF-2 high-resolution remote sensing images as the data source, and the Yesanpo National Park as the research area. It uses pixel-based MLC, NN, and SVM three classification methods to identify scenic resources, and adopts the system sampling method evaluation 3 methods to identify the classification accuracy of scenic resources, effectively improving the objectivity and accuracy of classification accuracy evaluation. The results show that the three classification methods used in GF-2 images meet the accuracy requirements of scenic resource identification, which is an effective method to identify scenic resources. Due to different geomorphic features, the erosion Zhanggu landform-Baili Xia scenic areas uses MLC classification, and granite fracture structure The canyon landform-LongmenTianguan scenic areas uses NN classification, and the karst cave spring landform-Yugu Dong scenic areas uses SVM classification with the highest overall accuracy. The results show that the proposed method can be applied to the identification of large-scale, high-resolution scenic resources, and provide more refined data support for scenic resource analysis and sustainable development of local tourism.