HJ-1A HSI Research Articles

Estimation of Soil Cations Based on Visible and Near-Infrared Spectroscopy and Machine Learning

Soil exchange cations are a basic indicator of soil quality and environmental clean-up potential. The accurate and efficient acquisition of information on soil cation content is of great importance for the monitoring of soil quality and pollution prevention. At present, few scholars focus on soil exchangeable cations using remote sensing technology. This study proposes a new method for estimating soil cation content using hyperspectral data. In particular, we introduce Boruta and successive projection (SPA) algorithms to screen feature variables, and we use Guangdong Province, China, as the study area. The backpropagation neural network (BPNN), genetic algorithm–based back propagation neural network (GABP) and random forest (RF) algorithms with 10-fold cross-validation are implemented to determine the most accurate model for soil cation (Ca2+, K+, Mg2+, and Na+) content estimations. The model and hyperspectral images are combined to perform the spatial mapping of soil Mg2+ and to obtain the spatial distribution information of images. The results show that Boruta was the optimal algorithm for determining the characteristic bands of soil Ca2+ and Na+, and SPA was the optimal algorithm for determining the characteristic bands of soil K+ and Mg2+. The most accurate estimation models for soil Ca2+, K+, Mg2+, and Na+ contents were Boruta-RF, SPA-GABP, SPA-RF and Boruta-RF, respectively. The estimation effect of soil Mg2+ (R2 = 0.90, ratio of performance to interquartile range (RPIQ) = 3.84) was significantly better than the other three elements (Ca2+: R2 = 0.83, RPIQ = 2.47; K+: R2 = 0.83, RPIQ = 2.58; Na+: R2 = 0.85, RPIQ = 2.63). Moreover, the SPA-RF method combined with HJ-1A HSI images was selected for the spatial mapping of soil Mg2+ content with an R2 of 0.71 and RPIQ of 2.05. This indicates the ability of the SPA-RF method to retrieve soil Mg2+ content at the regional scale.

Estimation of Cultivated Land Quality Based on Soil Hyperspectral Data

Efficient monitoring of cultivated land quality (CLQ) plays a significant role in cultivated land protection. Soil spectral data can reflect the state of cultivated land. However, most studies have used crop spectral information to estimate CLQ, and there is little research on using soil spectral data for this purpose. In this study, soil hyperspectral data were utilized for the first time to evaluate CLQ. We obtained the optimal spectral variables from dry soil spectral data using a gradient boosting decision tree (GBDT) algorithm combined with the variance inflation factor (VIF). Two estimation algorithms (partial least-squares regression (PLSR) and back-propagation neural network (BPNN)) with 10-fold cross-validation were employed to develop the relationship model between the optimal spectral variables and CLQ. The optimal algorithms were determined by the degree of fit (determination coefficient, R2). In order to estimate CLQ at the regional scale, HuanJing-1A Hyperspectral Imager (HJ-1A HSI) data were transformed into dry soil spectral data using the linkage model of original soil spectral reflectance to dry soil spectral reflectance. This study was conducted in the Guangdong Province, China and the Conghua district within the same province. The results showed the following: (1) the optimal spectral variables selected from the dry soil spectral variables were 478 nm, 502 nm, 614 nm, 872 nm, 966 nm, 1007 nm, and 1796 nm. (2) The BPNN was the optimal model, with an R2(C) of 0.71 and a normalized root mean square error (NRMSE) of 12.20%. (3) The results showed the R2 of the regional-scale CLQ estimation based on the proposed method was 0.05 higher, and the NRMSE was 0.92% lower than that of the CLQ map obtained using the traditional method. Additionally, the NRMSE of the regional-scale CLQ estimation base on dry soil spectral variables from HJ-1A HSI data was 2.00% lower than that of the model base on the original HJ-1A HSI data.

Estimation of Soil Nutrient Content Using Hyperspectral Data

Soil nutrients play a vital role in plant growth and thus the rapid acquisition of soil nutrient content is of great significance for agricultural sustainable development. Hyperspectral remote-sensing techniques allow for the quick monitoring of soil nutrients. However, at present, obtaining accurate estimates proves to be difficult due to the weak spectral features of soil nutrients and the low accuracy of soil nutrient estimation models. This study proposed a new method to improve soil nutrient estimation. Firstly, for obtaining characteristic variables, we employed partial least squares regression (PLSR) fit degree to select an optimal screening algorithm from three algorithms (Pearson correlation coefficient, PCC; least absolute shrinkage and selection operator, LASSO; and gradient boosting decision tree, GBDT). Secondly, linear (multi-linear regression, MLR; ridge regression, RR) and nonlinear (support vector machine, SVM; and back propagation neural network with genetic algorithm optimization, GABP) algorithms with 10-fold cross-validation were implemented to determine the most accurate model for estimating soil total nitrogen (TN), total phosphorus (TP), and total potassium (TK) contents. Finally, the new method was used to map the soil TK content at a regional scale using the soil component spectral variables retrieved by the fully constrained least squares (FCLS) method based on an image from the HuanJing-1A Hyperspectral Imager (HJ-1A HSI) of the Conghua District of Guangzhou, China. The results identified the GBDT-GABP was observed as the most accurate estimation method of soil TN ( of 0.69, the root mean square error of cross-validation (RMSECV) of 0.35 g kg−1 and ratio of performance to interquartile range (RPIQ) of 2.03) and TP ( of 0.73, RMSECV of 0.30 g kg−1 and RPIQ = 2.10), and the LASSO-GABP proved to be optimal for soil TK estimations ( of 0.82, RMSECV of 3.39 g kg−1 and RPIQ = 3.57). Additionally, the highly accurate LASSO-GABP-estimated soil TK (R2 = 0.79) reveals the feasibility of the LASSO-GABP method to retrieve soil TK content at the regional scale.

Prediction of Soil Nutrient Contents Using Visible and Near-Infrared Reflectance Spectroscopy

Quickly and efficiently monitoring soil nutrient contents using remote sensing technology is of great significance for farmland soil productivity, food security and sustainable agricultural development. Current research has been conducted to estimate and map soil nutrient contents in large areas using hyper-spectral techniques, however, it is difficult to obtain accurate estimates. In order to improve the estimation accuracy of soil nutrient contents, we introduced a GA-BPNN method, which combined a back propagation neural network (BPNN) with the genetic algorithm optimization (GA). This study was conducted in Guangdong, China, based on soil nutrient contents and hyperspectral data. The prediction accuracies from a partial least squares regression (PLSR), BPNN and GA-BPNN were compared using field observations. The results showed that (1) Among three methods, the GA-BPNN provided the most accurate estimates of soil total nitrogen (TN), total phosphorus (TP) and total potassium (TK) contents; (2) Compared with the BPNN models, the GA-BPNN models significantly improved the estimation accuracies of the soil nutrient contents by decreasing the relative root mean square error (RRMSE) values by 15.9%, 5.6% and 20.2% at the sample point level, and 20.1%, 16.5% and 47.1% at the regional scale for TN, TP and TK, respectively. This indicated that by optimizing the parameters of BPNN, the GA-BPNN provided greater potential to improving the estimation; and (3) Soil TK content could be more accurately mapped by the GA-BPNN method using HuanJing-1A Hyperspectral Imager (HJ-1A HSI) (manufacturer: China Aerospace Science and Technology Corporation; Beijing, China) data with a RRMSE value of 20.37% than the soil TN and TP with the RRMSE values of 40.41% and 34.71%, respectively. This implied that the GA-BPNN model provided the potential to map the soil TK content for the large area. The research results provided an important reference for high-accuracy prediction of soil nutrient contents.

Fusion Method Evaluation and Classification Suitability Study of Wetland Satellite Imagery

Based on HJ-1A HSI data and Landsat-8 OLI data, RS image fusion experiments were carried out using three fusion methods: principal component (PC) transform, Gram Schimdt (GS) transform and nearest neighbor diffusion (NND) algorithm. Four evaluation indexes, namely mean, standard deviation, information entropy and average gradient, were selected to evaluate the fusion results from the aspects of image brightness, clarity and information content. Wetland vegetation was classified by spectral angle mapping (SAM) to find a suitable fusion method for wetland vegetation information extraction. The results show that PC fusion image contains the largest amount of information, GS fusion image has certain advantages in brightness and clarity maintenance, and NND fusion method can retain the spectral characteristics of the image to the maximum extent; Among the three fusion methods, PC transform is the most suitable for wetland information extraction. It can retain more spectral information while improving spatial resolution, with classification accuracy of 89.24% and Kappa coefficient of 0.86.

Quantitative Inversion of Vegetation Biochemical Components Based on HJ1-A HSI in Coal Mining Area

This study investigates the applicability of estimating chlorophyll and water content at canopy level through empirical models and band combinations. The main goal is to evaluate and compare the accuracy of these two approaches for estimating and mapping canopy chlorophyll and water content through canopy reflectance and spaceborne HJ1-A HSI data acquired over Yanzhou coal mining area. An experiment was carried out. Canopy spectral measurements were acquired in the field using an ASD spectroradiometer along with simultaneous in situ measurements of leaf chlorophyll content. We tested seven variables derived from canopy reflectance for detecting canopy chlorophyll and water content: (1) R, (2) Log(1/R), (3) Log(1/R)′, (4) FDR, (5) SDR, (6) CRR, (7) BD. Stepwise multiple linear regressions were used to select wavelengths from HJ1-A HSI image bands. Correlation analysis was also done between different band combinations and biochemistry. A statistically significant relationship between Log(1/R) and chlorophyll was found at canopy level (R2 = 0.516). SDR had the highest correlation with canopy water content (R2 = 0.490). In addition, relationship between normalized different band combinations and chlorophyll and water content is also significantly obvious (R2 = 0.577 and R2 = 0.615). Canopy chlorophyll content was estimated with the intermediate accuracy (R2 = 0.4144), while water content was estimated with an acceptable accuracy (R2 = 0.4592). Canopy chlorophyll and water content spatial distribution were mapped. Chlorophyll and water stress levels were quantified by comparing different environmental stressors.

A Subspace-Based Change Detection Method for Hyperspectral Images

Remote sensing change detection has played an important role in many applications. Most traditional change detection methods deal with single-band or multispectral remote sensing images. Hyperspectral remote sensing images offer more detailed information on spectral changes so as to present promising change detection performance. The challenge is how to take advantage of the spectral information at such a high dimension. In this paper, we propose a subspace-based change detection (SCD) method for hyperspectral images. Instead of dealing with band-wise changes, the proposed method measures spectral changes. SCD regards the observed pixel in the image of Time 2 as target and constructs the background subspace using the corresponding pixel in the image of Time 1, and additional information. In this paper, two types of additional information, i.e., spatial information in the neighborhood of the corresponding pixel in Time 1, and the spectral information of undesired land-cover types, are used to construct the background subspace for special applications. The subspace distance is calculated to determine whether the target is anomalous with respect to the background subspace. The anomalous pixels are considered as changes. Here, orthogonal subspace projection is employed to calculate the subspace distance, which makes full use of the advantage of the abundant spectral information in hyperspectral imagery, and is also easy to apply. The experimental results using Hyperion data and HJ-1A HSI data indicate that SCD gives more accurate detection results, with a lower false alarm rate, compared with other state-of-the-art methods. SCD with additional information also gives satisfactory results in the experiments, reducing the false alarms caused by misregistration and suppressing the change of undesired land-cover types.

Using Data of HJ1A HSI for Retrieving Grassland Leaf Area Index

Leaf area index(LAI) is one of the key parameters for quantifying the energy and mass exchange characteristics of terrestrial ecosystems.To validate the feasibility of using HJ-1A HSI data to retrieve leaf area index,HJ-1A HSI data was acquired from China Center for Resource Satellite Data and Applications while in situ LAI was collected with a LAI-2000 sensor.Based on the field experiment data,an evaluation of grass LAI retrieval methods was conducted using three vegetation indices including ratio vegetation index(RVI),normalized difference vegetation index(NDVI) and modified soil adjusted vegetation index(MSAVI).The results showed that the three vegetation indices derived from HSI data have good relationship with measured LAI.MSAVI gained the best result with R2 value of 0.622.While the cross validation result of MSAVI model was also permissible with R2 value of 0.547 and RMSE value of 0.202.All the above showed that the measured LAI and simulated LAI have good consistency.At last,the map of grass LAI,which could partially reflect the growing status of the pasture,was produced using HJ-1A HSI data and MSAVI model.

基于HJ 星高光谱数据红边参数的冬小麦叶面积指数反演

This study intends to examine the applicability of HJ-1A HSI data to estimation of leaf area index for crop monitoring. Four red edge indices (red vale position, red edge position, red edge slop and red edge swing) were used to extract the required information from the HJ-1A HSI data with inverse Gauss model. Combined with ground measurements data in April and May in 2009, regression analyses were used to establish red edge parameters model for LAI. Red edge position, red edge slop and red edge swing were highly correlated with LAI. R 2 were 0.5592, 0.7796 and 0.8107 respectively. Validation results showed that the HJ-1A HSI data have great potential in LAI inversion.

Research on Standard Preprocessing Flow for HJ-1A HSI Level 2 Data Product

To deal with the Level 2 HSI data from the newly-launched HJ-1A satellite,this paper introduced in details the entire flow and relevant algorithms for data preprocessing.The introduction includes calibration,vertical stripes elimination,and atmospheric correction geometric correction.Standard spectral reflectance products with precise geo-locations were produced.Spectral reflectance data from EO-1 Hyperion of close dates was used to simulate the band reflectance of HJ-1A HSI.Comparisons of spectral reflectance data between simulated and actual HJ-1A HSI were made to validate the effect of the data preprocessing.The average correlation coefficient of spectral reflectance between actual and simulated HJ-1A HSI is 0.947 with its standard deviation being 0.017,suggesting a high consistency.The mean and standard deviation of differential bands between real and simulated HJ-1A HSI are close to 0.The result shows that the reflectance from HJ-1A HSI is consistent with that of simulated data from Hyperion,and hence the data processing flow could provide necessary support for quantitative use of HJ-1A HSI data.