最新刊期
卷 24 ， 期 3 ， 2020
-
摘要：GF-4 is China's first geostationary orbit optical remote sensing satellite with high-resolution ground observation. Compared with other meteorological satellites, it has a single medium-wave infrared channel with the characteristics of high spatial and temporal resolution. To explore the application of GF-4 panchromatic multispectral and medium wave infrared (PMI) in forest fire detection and provide a new method for forest fire monitoring in China.Yulong autonomous county of Yunnan province, the Amur region and the outer Baikal frontier of Russia were selected as experimental areas. During the forest fires in the experimental areas in 2017 and 2018, 12 scene images of GF-4 PMI were obtained, of which 8 scene images were used as the experimental group, and 4 scene images were used as the verification group. Statistical analysis was performed on the typical characteristics of the experimental group images, and the ‘split window method’ was used to construct an adaptive threshold detection algorithm, then the detection algorithm was used to detect the images of the verification group and compared with the results by visually interpreted.The results showed that the detection of forest fire points in Yulong autonomous county of Yunnan province, the accuracy of the algorithm in this paper was 80.0%, the omission detection rate was 20.0%, and the comprehensive evaluation index, which was based on the verification of fire detection, was 0.781. The detection of forest fire points in the outer Baikal frontier of Russia, the accuracy of the algorithm detection was 99.1%, the omission detection rate was 24.3%, and the comprehensive evaluation index was 0.858. The forest fire of 2017 in the Amur region, the accuracy of the algorithm detection was 97.7%, the omission detection rate was 22.2%, and the comprehensive evaluation index was 0.866. The forest fire of 2018 in the Amur region, the accuracy of the algorithm detection was 92.4%, the omission detection rate was 14.5%, and the comprehensive evaluation index was 0.889.The accuracy of the fire points detection in these three experimental areas was higher than 80.0%, the comprehensive evaluation index set based on the accuracy verification of the fire point detection was higher than 0.780. This algorithm could realize the fire point detection of GF-4 PMI images, and the algorithm had a higher accuracy rate of fire point detection in a large range of forest fire, but the omission detection rate of the algorithm was high and needs to be further optimized. The experimental results showed that the proposed algorithm was reliable, which could provide a method reference for forest fire monitoring in China.关键词：GF-4;split-window method;adaptive method;fire point;threshold detection- 670
- |
- 131
- |
- 6
发布时间：2021-10-18 -
摘要：The Earth’s radiation was observed by a microwave radiation imager (MWRI) on board FengYun-3D (FY-3D) satellite at 10.65 GHz, 18.7 GHz, 23.8 GHz, 36.5 GHz, and 89.0 GHz with dual polarization. The nonlinearity of this payload, as an important parameter in calibration algorithms, is represented by the nonlinear parameter- 588
- |
- 101
- |
- 1
发布时间：2021-10-18 -
摘要：Based on characteristic of High Spatial Resolution (HSR) image, an adaptive local geometric invariant feature extraction method based on Gabor transform and Tensor Voting (TV) is proposed and applied to building area extraction from HSR remote sensing images. First, image geometric features are analyzed. Second, the feasibility of extracting built-up area using local feature points and probability density estimation are determined. This study provides specific methods and steps. First, considering the abundant geometric features of high spatial resolution remote sensing images, multi-scale and multi-direction Gabor filter banks are adopted in detecting the singularity of images contain building area. In order to extract edge information of buildings, only real part Gabor coefficients are used. Meanwhile, we also measured the influence of Gabor kernels of different size on geometric feature extraction experiments, and the optimal scale parameter intervals for geometric feature extraction of remote sensing images with different spatial resolutions are thus determined. Second, it is a fact that the response of Gabor filter at each pixel is a measure for orientation certainty, thus, we introduce the orientation tensor which represents an ideal direction in the direction of the unit vector perpendicular to frequency. By weighting the orientation tensor with Gabor coefficients ans summing over it to complete the information fusion, the resulting tensor givens an estimate for local orientation and orientation uncertainty at image position. The tensors are then used as an initial estimate for global context refinement using tensor voting and the points are classified based on the their likelihoods of being part of a feature type, non-maximal suppression is used to extract geometric features. A key advantage of combining the Gabor filtering and tensor voting is that it eliminates the need for any thresholds therefore removing any data dependencies. In order to achieve a reliable extraction of local invariant feature such as corners from built-up area, three criterions are further proposed to refine the first stage result. Finally, each local building corner indicates a building be detected in image. However, only one of them is not sufficient alone to detecting a building. In fact, the more local points a buildings has, the more probable its detection becomes. Based on that fact, a probability density estimate method is generated to describe the probability that the pixel belongs to the building area, and the Otsu method is used to automatically extract the polygon area of the residential area. Experiments were carried out using image data sets such as Google and GF-2 with resolution higher than 1 meter, results showing that the proposed method can achieve higher accuracy in building area detection compared with the state-of-the-art corner detection algorithm such as Harris corner and High-speed corner detect method.关键词：high spatial resolution remote sensing image;geometric feature extraction;built-up area detection;Gabor transform;tensor voting;probability density field;GF-2- 616
- |
- 165
- |
- 6
发布时间：2021-10-18
Chinese Satellites
-
摘要：Multiscale analysis technique can describe an image from different resolutions and has been widely used for extracting features and modelling of remotely sensed images. The subsampled wavelet transforms are commonly employed for establishing the multiscale representation of an image. However, the wavelet-based features cannot describe patterns with long spatial spans and often result in noisy classification results. By contrast, the object-based image analysis can create classification maps composed of compact land objects. However, features extracted from a single scale still cannot provide discriminating information for the land cover classification. To improve the classification accuracy and alleviate noisy thematic maps, a regional multiscale classification method is proposed.The proposed method consists of two main blocks, including establishing a regional multiscale image representation and classifying on the basis of Markov Random Field (MRF). In the first block, the mean shift segmentation method is employed to create initial over-segmented regions. Thereafter, a rule combining the grey values in the regions and the shared boundary lengths among regions is designed to extract the low-frequency part of an image. The current high-frequency part is obtained by subtracting the original image and the current low-frequency portion. A regional multiscale representation can be iteratively established by replacing the original image with the low-frequency part and repeating the segmenting and decomposing process. In the second block, the classification result obtained from the original image is considered the prior of the label field in the first decomposed level, and the high-frequency part in the first level models the feature field of MRF. The classification result in the first level is obtained by solving an objective function consisting of feature and label energies. By iteratively projecting the current classification result to the next level and modelling the feature field with the high-frequency part, the final classification map can be acquired in the coarsest scale.To examine the effectiveness of the proposed method, three groups (six images in total), such as Prague textures, synthetic remote sensing textures, and QuickBird multispectral images, are adopted. The proposed regional multiscale MRF (RMRF) is compared with the Iterated Conditional Model (ICM), Graph Cut (GC), and Wavelet-based MRF classification methods (WMRF). Visual inspection and quantitative measurement are employed for the comparison. As GC fails to integrate the spatial information among neighbouring pixels, the classification results are dominated by the pixel spectral values. Textures or land cover patterns with complex spectral heterogeneity cannot be properly captured. WMRF provides a pixel-level multiscale representation, which is also insufficient for describing texture patterns across a large spatial span. ICM, GC and WMRF may create noisy thematic maps with different extensions when the spectral heterogeneity of a given land cover pattern is extreme. On the contrary, RMRF has a flexible framework to extract and model spatial information. RMF can capture the essential features of land objects with large spectral heterogeneity, resulting in maps with less noise.This study proposes an RMRF for classification. Experiments demonstrate that the established multiscale representation can efficiently describe texture patterns with wide spatial spans, such as land objects with complex structures. By combining the multiscale representation with MRF model, RMRF can achieve a high-precision semantic segmentation of ground objects. The adaptive estimation of the decomposition layer number and parameters of the algorithm are the on-going works.关键词：high resolution remote sensing;image classifications;regional segmentation;Markov random field;multiscale model- 576
- |
- 184
- |
- 8
发布时间：2021-10-18 -
摘要：Image semantic segmentation refers to segmenting an image into several groups of pixel regions with different specific semantic meanings and identifying the categories of each region. In recent years, the common semantic segmentation methods that are based on Convolutional Neural Networks (CNN) have realised the pixel-to-pixel image semantic segmentation. They can avoid the problems of artificial design and selection of features in traditional image semantic segmentation methods. As a result of the pooling operation and lack of context information, the detailed information of images is neglected, the precision of the final image semantic segmentation result is low and the segmentation edge is inaccurate. Therefore, this study proposes a semantic segmentation method for remote sensing image on the basis of Deep Fusion Networks (DFN) combined with a conditional random field model.The method initially builds a DFN model in a Fully Convolutional Network (FCN) framework with a deconvolutional fusion structure. On the one hand, the multiscale features can be extracted through the deep networks, which can avoid the artificial design and selection of features to improve the generalisation ability of the model. On the other hand, the multiscale information is used in the model with the help of the deconvolutional fusion structure. The processing accuracy of the model is also improved by fusing the shallow detail information and deep semantic information. Fundamentally, the fully connected conditional random field is introduced to supplement the spatial context information towards precisely locating the boundary and obtaining final semantic segmentation results.From this study, we can draw the following(1)With the increase in the depth of the fusion layer, detailed information becomes abundant, the semantic segmentation results become refined and the edge contour becomes close to the label image;(2) The fully connected conditional random field model synthesises the global and local information of the remote sensing image and further improves the efficiency and accuracy of the final semantic segmentation results.关键词：remote sensing image semantic segmentation;fully convolutional networks;conditional random field;fusion structure;deconvolution- 738
- |
- 279
- |
- 5
发布时间：2021-10-18 -
摘要：Geothermal exploration in mountain regions strongly relies on the successful identification of thermal anomaly in the regions, which can be performed through the extraction of geothermal anomaly information from thermal infrared remote sensing data. However, changes in Land Surface Temperature (LST) in mountain regions are significantly affected by topography in addition to other factors, such as latitude and surface property. This effect strongly weakens the efficient identification of geothermal anomaly over the LST image retrieved from thermal infrared remote sensing data, which consequently limit the application of remote sensing technique to geothermal resource exploration in mountain regions with rough terrain. This study examines the effects of terrain on LST changes in the mountain region of Longchuan in Southern China to establish an efficient approach, which can correct the effects of terrain on the LST changes for the geothermal exploration in the region.LST was retrieved using mono-window algorithm based on Landsat ETM+ remote sensing data. The effects of terrain on LST distribution were then analysed. The statistical analysis showed a parabolic relationship between LST and aspect. Moreover, the southeast-facing slope had the highest average LST and standard deviation, where LST was significantly and positively correlated with slope gradient. To reduce the impact of terrain on LST distribution, the area was divided into sunny slope, shady slope and transitional slope. The LST in the sunny slope was particularly corrected to its horizontal surface equivalent by the linear regression equation between LST and slope gradient. Geothermal anomalies were then extracted from the LST of these three subareas, with the consideration of geologic structure and land cover.Results showed that the spatial variation amplitude of LST evidently decreased because the significant temperature difference among different terrain conditions has become small in subareas. Four possible geothermal anomalies were recognised in which high temperature areas were closely related to faults and showed little variability in land cover. Comparative analysis with known hot springs indicated that they were likely caused by geothermal activities.In conclusion, topography mainly affects LST spatial distribution by controlling incoming solar radiation. The solution of aspect-based partition and gradient correction presented in this article can also effectively reduce topographic effects. It helps improve the recognition accuracy of geothermal anomalies with remote sensing technology. The solution may also provide an enlightening insight into the forecast evaluation of geothermal resources in mountain regions. Moreover, further analyses of the relationship between LST and other factors controlled by terrain are necessary in future research, especially the physical properties of underlying surfaces, such as land use, soil moisture and vegetation.关键词：geothermal resource;land surface temperature;terrain effect;Landsat ETM+;thermal infrared;aspect;anomaly extraction- 654
- |
- 184
- |
- 3
发布时间：2021-10-18
Technology and Methodology
-
摘要：Hyperspectral Compressed Sensing (HCS) is crucial for data storage and the real-time transmission of airborne- or spaceborne-based imaging platforms. The Linear Mixing Model (LMM) has been successfully applied to HCS reconstruction. However, the obtained spectrum may be disturbed, thereby limiting the improvement of reconstruction quality due to the influence of illumination conditions, topographic changes, and atmospheric effects. Spectral disturbance is corrected on the basis of LMM by introducing the spectral correction term, and a linear mixing model for spectral perturbation correction is proposed. Moreover, an improved HCS method based on modified LMM is proposed. This proposed model only performs spectral compressed sampling on the original hyperspectral images at the sampling end. The proposed method uses the proposed spectral perturbation correction model to reconstruct the original hyperspectral images based on the compressed sampling data. The alternating direction multiplier method is used to estimate the optimal values of each component in the modified LMM to obtain the optimal reconstruction quality. Experimental results show that the proposed method can achieve better reconstructed performance compared with other classical HCS methods.关键词：remote sensing;hyperspectral remote sensing;compressed sensing;linear mixing model;spectral perturbed- 571
- |
- 201
- |
- 1
发布时间：2021-10-18 -
摘要：One of the greatest advantages of microwave remote sensing over other remote sensing techniques is penetrability. Quantitatively estimating the sensing depth of passive microwave remote sensing is meaningful for simulation of satellite signals and validation of land surface parameters to estimate the sensing depth of passive microwave remote sensing. In this paper, a simple statistical model for estimating the thermal sampling depth in microwave frequencies of soil was developed and validated.Thermal Sampling Depth (TSD) was introduced to describe the source of the main signals of passive microwave remote sensing. To develop a simple statistical model for estimating the TSD of soil, a theoretical model was introduced to describe the emission characteristics of a three-layer case, which incorporates all multiple reflections at the two boundaries. Based on radiative transfer theory, the total emission of the three layers was calculated. Sensitivity analysis was then performed to demonstrate the effects of the soil properties and frequency on the TSD based on a simulation database covering a wide range of soil characteristics and frequencies. Based on the sensitivity analysis results, a statistical model for estimating TSD was developed. This model can estimate the TSD using four common and easily acquired parameters: soil moisture, temperature, frequency, and soil specific surface area. For validation, a controlled field experiment using a Truck-mounted Multi-frequency Microwave Radiometer (TMMR) was designed and performed.The total Root Mean Square Error (RMSE) between the TSD measured in field experiment and estimated using the statistical model was approximately 0.5 cm for the TRMM’s four frequencies.The results indicated that the developed statistical model offers a relatively accurate and simple way to estimate the TSD.关键词：passive microwave remote sensing;thermal sampling depth;statistical model;soil moisture;validation- 546
- |
- 162
- |
- 0
发布时间：2021-10-18
Technologies and Methodologies
-
摘要：This study aims to analyze the vertical distribution of aerosol particles during autumn and winter by deriving the number density of aerosol particles from Mie Lidar return signals. Haze has always existed in regions with low humidity and stable weather. In such regions, particle vertical distribution characters are needed for transport mechanisms. As a widely used instrument, Mie Lidar surfer from the uncertainties caused by the Lidar ratio and boundary value assumption. However, Mie Lidar is the most powerful tool to observe the atmospheric vertical distribution because it performs well in day and night under all weather conditions in comparison with passive sensors. Moreover, Mie Lidar is easier to carry compared with Raman Lidar.Lidar radiation principles were studied to understand the Lidar equation and its solution. Several parameters were carefully chosen for the numerical solutions. Instead of using boundary value for approximation, the Lidar parameter k was directly calculated on the basis of Lidar-validated parameters to avoid uncertainties. Thus, aerosol particles backscatter and extinction coefficient were retrieved by solving the Lidar equation and integrated for Aerosol Optical Depth (AOD) results. A Cimel CE318 sun photometer and an ASD FieldSpec3 were used to measure AOD as validation data. Lastly, the number density of aerosol particles was derived on the basis of Mie theory with the estimation of aerosol particle distribution, size, and refractive index of particles.According to the number density vertical profile, nearly no difference was observed between the number density vertical profile of aerosol particles during daytime and nighttime in one day at a low altitude. This result indicated that the sunlight effect is ignorable for aerosol particle distribution near ground. The retrieved AODs with the CE318 and ASD observations showed a root mean square error of approximately 0.0541 and 0.0100, demonstrating that the retrieved result is reliable. The comparison with traditional method showed that the proposed algorithm in this study improved the accuracy of the traditional Lidar equation solution. The near-real distribution results showed aerosol particle distribution characteristics near the ground. The number density of aerosol particles in Guangzhou was mainly distributed in low-altitude regions, and its relationship with height was approximately a negative exponential distribution. The retrieved results of the aerosol particle number density were similar to those from the previous years, indicating that the air pollution in Guangzhou has not worsened. However, potential environmental threats and air pollution problems still need to be governed.Several uncertainties, such as changes in air temperature and pressure, validation of Lidar instrument k, and overlap factor, that could cause deviation were identified. The overlap factor and molecular backscatter coefficient were carefully calculated using the theoretical method. Frequent updates on Lidar parameters will be requested from the Lidar company in the future. The theoretical and experimental methods were combined for the calculation of the overlap factor and molecular backscatter coefficient. Raman Lidar was used to measure the actual Lidar ratios and characteristics of molecular scatter.The aerosol particle number density-retrieved algorithm was based on Lidar radiation principles and revealed the particles’ spatial distribution characters. The vertical distribution characters remained unchanged for several days under stable weather in the autumn and winter seasons, and its relationship with height was approximately a negative exponential distribution; hence, air conditions can have huge effect on human health. Air pollution problem still needs to be governed.关键词：Mie Lidar;Lidar radiation principles;aerosol;aerosol particles;mass concentration;Pearl River Delta- 598
- |
- 165
- |
- 2
发布时间：2021-10-18 -
摘要：Atmospheric CO2 concentration over China has significant effects on the global climate change. To reliably predict the impact of atmospheric CO2 on global climate change, it is necessary to clarify the distribution and variation of atmospheric CO2 concentration. Based on long term short-wavelength infrared CO2 dataset observed by GOSAT, the temporal variation and spatial distribution characteristics and variation trend of atmospheric CO2 concentration was investigated and analysed over China during 2010 to 2016. To ensure the quality of GOSAT CO2 dataset used in this paper, the GOSAT XCO2 dataset was validated with high precise XCO2 from ground-based TCCON sites. Multi-year mean of XCO2 was illustrated to show the spatial heterogeneity of CO2 concentration over China. Interannual variation and annual growth of XCO2 was also presented and discussed. The results showed that GOSAT XCO2 dataset was biased by -1.04±2.10 ppm with respect to TCCON XCO2, and the correlation coefficient was 0.90 between them. Seven years (2010~2016) of GOSAT CO2 dataset showed that high CO2 concentrations were mainly located in Zhejiang-Jiangsu-Anhui region, Beijing-Tianjin-Hebei region, and Hunan-Hubei-Henan-Shanxi region in China. The CO2 concentration reached 400 ppm over most regions in China until 2016. The annual average of CO2 concentration showed an increase trend year by year over China, increasing from 387.76 in 2010 to 402.18 ppm in 2016. The annual growth rate of CO2 concentration was evaluated to be 2.31 ppm/a during this period over China, which was slightly higher than the average in the world. This paper shows that the CO2 concentration observations from satellites could provide some references for the climate change response strategies and atmospheric environment control.关键词：CO2 concentration;distribution pattern;annual growth rate of CO2;China mainland;GOSAT- 834
- |
- 311
- |
- 14
发布时间：2021-10-18
Remote Sensing Applications
-
摘要：The Desert Locust (Schistocercagregaria) has been an important agricultural pest at least since biblical times. Desert Locust Swarms in Northeast Africa and the India-Pakistan border swept through many countries at the end of 2019 and in early 2020. Large areas of farmland and natural vegetation were eaten, threatening local agricultural and animal husbandry production. The occurrence and development of Desert Locusts of African is closely related to climatic factors such as local precipitation (soil moisture), temperature, wind speed and wind direction. So what climatic conditions contributed to this Desert Locust Swarms? And the Locust Swarms near the India-Pakistan border where is the closest Locust Swarms to China has become a hot spot for research and Chinese media. How did the Locust Swarms near the India-Pakistan border effect on the local vegetation? What is its development trend? Is it possible to fly into China? The impact of Desert locusts of the Indian-Pakistani border on local plants was analyzed by using time series satellite remote sensing data. The climatic factors connection with takeoff and migration of the Desert Locust Swarms were summarized based on past literatures, and the expansion trend of the Desert Locust Swarms and its possibility of entering China were analyzed based on the climatic characteristics such as precipitation and temperature in Africa and West Asia. It was concluded that: (1) due to the vegetation gnawing by the Desert Locust Swarms, the Normalized Difference Vegetation Index(NDVI) of the large-scale area decreased obviously compared with normal years in January and February 2020 in Indian-Pakistani border area, and the area of NDVI decreased was enlarged in February compared to January; (2) Several rare cyclones(two in 2018,one in 2019) that brought strong precipitation to eastern Africa and the Arabian Peninsula played an important role for this Africa horn-West Asia locust plague; (3) After analyzing the swarms takeoff temperature and Desert Locusts suitable breeding moisture( precipitation conditions), it suggested that the Locust Swarms had rare chance to migrate to eastern India, and it is even less likely to enter China.关键词：Desert locust;india-pakistan border;satellite remote sensing data;meteorological observation data;tropical storm;Normalized Difference Vegetation Index (NDVI)- 642
- |
- 120
- |
- 3
发布时间：2021-10-18
Short Communications
- Postal code：100190
- Tel：010-58887052 Email：nrsb@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备20021838号-9
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0