The concentric cable-driven manipulator has characteristics of flexibility and slenderness when working in confined space. However, how to optimize a structure with appropriate stiffness becomes an important issue. In this paper, a structure optimization method based on stiffness model is proposed for the designed concentric cable-driven manipulator with three segments. Firstly, according to the characteristics of the cable and center support, the concentric cable-driven manipulator with three segments is designed in this paper. The concentric cable-driven manipulator addresses flexibility and active regulation of cable tension. Secondly, the stiffness modeling of the concentric cable-driven manipulator with three segments is carried out by introducing the concept of unit module. The stiffness of the concentric cable-driven manipulator is adjusted by optimizing cable diameters. Finally, the stiffness of C-shaped and S-shaped configurations are analyzed and compared. The analysis and experimental results can provide reference for structural optimization in practical applications.

This article presents a segment-progress statics method for configuration prediction or force sensing of concentric cable-driven manipulators. First, the statics model of the whole system was constructed by progressively analyzing from local segments to the whole. This statics modeling method reduces the complexity of the model and considers the most common forces and moments. Second, a solving flow is proposed based on the iteration-correction strategy for configuration prediction or force sensing. This strategy provides a new idea for the analysis of the static friction direction when the statics model is used for quasi-static prediction. Simulations and experiments show that the proposed segment-progress statics method is effective for configuration prediction or force sensing of concentric cable-driven manipulators. Moreover, the effect of the friction gain on the accuracy of statics model is found by analyzing the curvature of the manipulator. The piecewise constant curvature (PCC) assumption is also concluded to be only applicable to small end forces but not to large end forces for concentric cable-driven manipulators.

Concentric cable-driven manipulators have the flexibility and typical advantages of working in confined environments. However, its configuration optimization in confined three-dimensional (3-D) space is very complicated due to infinite configurations of inverse kinematics solutions. This article proposes a spatial triarc planning method in response to the issue mentioned above. A reasonable triarc configuration can be optimized by this method based on six input parameters, i.e., the proximal control point and tangent vector, distal control point and tangent vector, and two centers of curvature circles in both two-dimensional (2-D) and 3-D space. This method has the following three advantages. First, the task configuration of a spatial triarc can be predicted by presetting the relationship between the proximal and distal tangent vectors. Furthermore, the configuration of the middle and inner concentric cable-driven mechanism can be controlled by adjusting the direction of the distal tangent vector. Additionally, the proportion of each concentric cable-driven mechanism can be controlled by changing centers of curvature circles. Finally, the proposed spatial triarc planning method is verified by simulations and experiments. Results show that the maximum errors of the C-shaped spatial triarc and the S-shaped spatial triarc experiments are 1.68 mm and 1.36 mm, respectively.

Low-temperature disasters seriously affect the ecological environment and people's daily lives. In the research, the spatial distribution of low-temperature risk over 30 years has been obtained by collecting meteorological and socio-economic data from 1991 to 2020, and then a risk assessment model based on hazard, exposure, and vulnerability indicators is built. The results show that the high-risk areas are mainly concentrated in the central, eastern, and southern parts of Northeast China, and the overall risk shows an increasing trend; the regions at high risk are primarily found in Hunan Province and the central part of Jiangxi Province in south China. Further, the change in risk is most obvious between 1991-2005, and the overall risk value decreased about 41.28%. The Contribution to low-temperature risk in Northeast China comes mainly from vulnerability. While, In South China, the contribution of various factors to risk is more evenly distributed.

D-LinkNet has been a popular convolutional neural network. However, the scale of the model would be huge. Focused on the problem, this article made some improvements:(1) Replace the initial block with a stem block. (2) Add bottleneck layers before and after DBlock to form a new module called DBlockPlus. (3) Rebuild the entire network based on ResNet. The article also proposed a method named ESIPs to eliminate small independent patches and adopted a composite loss function called DF loss to better address the impact of extreme class imbalance in binary classification. And another composite loss function was introduced for comparative experiments. The experimental results show that the SBD-LinkNet not only has less parameters but also has a higher accuracy and the road extraction results presented a better visual effect after processing by ESIPs. And the model that uses DF loss also performs better in road extraction from complicated scene.

Tobacco is a critical cash crop in China, so its grow status has been attributed more and more attention. How to acquire the accurate plant area, row spacing and plant spacing have been the key points for its grow status monitoring and yield prediction. Currently, remote sensing has been a popular method for its speediness, large scale and costless, which could replace the traditional manual methods. We proposed a method to extract the planting information of tobacco at the rosette stage with UAV (Unmanned Aerial Vehicle) remote sensing images, and solved the following problems: the difficulty of detecting the small and densely planted tobacco objects, the scattered tobacco fields with different shapes in Sichuan Province. Four experimental areas were selected in Sichuan Province, and image processing and sample label production were carried out. The results indicate that the average accuracy of tobacco field area, row spacing and plant spacing extracted by this method reached 98.35%, 97.90% and 97.74%, respectively, which proved the extraction method of plant information are valuable.

Detecting semantic changes in high-resolution aerial images is a challenging task in remote sensing technology. Previous change detection methods have encountered difficulties in detecting subtle changes and have faced problems such as overfitting of image time series and unclear boundaries of the generated change maps. This paper proposes SEFNet, a multi-task change detection network based on symmetric enhancement fusion, to address the aforementioned issues. SEFNet is built on a decoupled three-branch change detection network framework, which allows each branch module to perform more specialized functions, facilitating more targeted training. Additionally, this study introduces the Symmetry-enhanced Fusion Change Detection unit (SFCD) module to extract multi-level change information. To highlight subtle changes while incorporating shallow detailed features of the network, the SFCD method employs feature enhancement methods and feature reuse strategies. Additionally, the proposed symmetric fusion module reduces the network's sensitivity to image time. A multitask classifier is also utilized to generate semantic change detection maps. The method's effectiveness is demonstrated on the SECOND dataset, achieving the best results in the current state-of-the-art dataset.

Semantic change detection (SCD) based on remote sensing images can provide an effective solution for large-scale land use monitoring. The existing change detection (CD) networks face limitations due to their limited receptive fields, which make it difficult to provide a comprehensive and consistent response to changes in similar large objects. Moreover, these limitations often cause the network to ignore weak changes localized in the images. To address these limitations, we propose change-guided similarity pyramid network (CG-SPNet), a decoupled multitask architecture that integrates three key components, including the multiscale positional similarity pyramid (MSSP), the symmetry-enhanced fusion CD unit (SFCD), and the change-guided feature interaction module (CGM). MSSP module captures positional correlation of semantic features at multiple scales by introducing an attention pyramid during pooling downsampling. SFCD utilizes a convolutional attention mechanism to enhance local features and highlight weak changes in dual images, and the symmetric fusion (SF) approach reduces the network’s sensitivity to timing. CGM is used to address the challenge of effective feature interaction, which combines cosine similarity loss to leverages cross-attention to compute the positional relevance of change features to embed a priori information for semantic features. Through rigorous experiments and analyses, we have successfully validated the essential components of CG-SPNet. It achieves the state-of-the-art performance on the SECOND dataset as well as our self-built CZWZ dataset.

This article studies the$H_\infty$exponential synchronization problem of delayed chaotic Lur'e systems (DCLSs). Considering the case of sensors sampling with different periods, an asynchronous memory-based event-triggered scheme is proposed to deal with the multiple sampling periods cases. To be specific, a group of memory-based event-triggered processors with different triggering conditions are set behind the sensors. Such a scheme supports the sensors to sample with different periods and supports the packets arrive at the controller side asynchronously. Then, to develop the closed-loop system, a merging time sequence$\lbrace t_{s}\rbrace$is constituted by using the release instants and by considering the transmission delays. On this basis, a so-called multirate Lyapunov functional is constructed, which including the information of the sampling upper bounds of different sensors. Furthermore, two criteria for$H_\infty$exponential synchronization of DCLSs are derived in the form of linear matrix inequalitys (LMIs). And, the controller gain can be obtained from the feasible solution of the LMIs. And, a numerical example is provided to demonstrate the effectiveness and merits of the proposed method.

This article investigates the sampled-data-based secure synchronization control problem for chaotic Lur'e systems subject to power-constrained denial-of-service (DoS) attacks, which can block data packets' transmission in communication channels. To eliminate the adverse effects, a resilient sampled data control scheme consisting of a secure controller and communication protocol is designed by considering the attack signals and periodic sampling mechanism simultaneously. Then, a novel index, i.e., the maximum anti-attack ratio, is proposed to measure the secure level. On this basis, a multi-interval-dependent functional is established for the resulting closed-loop system model. The main feature of the developed functional lies in that it can fully use the information of resilient sampling intervals and DoS attacks. In combination with the convex combination method, discrete-time Lyapunov theory, and some inequality estimate techniques, two sufficient conditions are, respectively, derived to achieve sampled-data-based secure synchronization of drive-response systems against DoS attacks. Compared with the existing Lyapunov functionals, the advantages of the proposed multi-interval-dependent functional are analyzed in detail. Finally, a synchronization example and an application to secure communication are provided to display the effectiveness and validity of the obtained results.

Semantic layout synthesis based on Generative Adversarial Networks (GANs) has made significant advances, however, the generation of high-quality real remote sensing images through semantic layouts remains a significant challenge. In this paper, we propose a Multi-Task Generative Adversarial Network (MTGAN) that utilizes semantic layouts to generate remote sensing images featuring five common geographic objects. Our approach takes into account the unique characteristics of remote sensing images, such as multiple scales, multiple objects, low inter-class separability, and high intra-class heterogeneity. The MTGAN utilizes a global-local GAN structure, with the global generator producing global context information and the local generator generating specific class information. By combining global macro information and local detail information, the MTGAN is able to generate remote sensing images with close contextual connections and clear details of geographical objects.

Tobacco is one of the most important cash crops in China, and accurate acquisition of its planting information is of great significance for tobacco management and yield prediction. In this paper, a tobacco information extraction method based on UAV high-resolution images is proposed. The method uses the RtinaNet target detection model to achieve the identification and localization of tobacco in UAV high-resolution images, and generates tobacco coordinate information. Based on the tobacco coordinates, we designed two algorithms for extracting plant spacing, row spacing and area of tobacco fields, respectively. The experimental results showed that the accuracy of the method reached 98.49%-99.77%, 97.08%-99.98% and 94.48%-99.63% for area, row spacing and plant spacing extracted from the seven experimental tobacco fields, respectively.

Soil plant analysis development (SPAD) represents relative chlorophyll content, which directly affects tabacco health. Accurate monitoring of tobacco canopy SPAD is vital to guide field management. Due to low correlation, few number of features and single model, the existed inversion models have low accuracy and poor robustness. This paper expanded samples from hyperspectral images using PROSAIL, so as to avoid overfitting. In the aspect of feature extraction, the optimized vegetation index is added to increase the correlation between features and SPAD. The inversion model combines K-means and XGBoost to form a mixed model. The results show that the mixed model has better effect on the validation set than other models, R2=0.83, RMSE=3.9.

In recent years, many semantic segmentation methods based on convolutional neural networks (CNN) have been applied to road extraction, but objects with similar spectral characteristics to roads in RGB images and road occlusions cause the discontinuous output of road extraction. To ensure extraction performance, hyperspectral data is used as a supplement to RGB data to improve the ability of remote sensing image road extraction in this paper. This paper uses a multi-scale fusion attention network to combine RGB and multispectral imagery. First, band selection is used to select multispectral bands with high inter-class separability, and the Cross-Source Feature Recalibration Module (CSFR) is used to calibrate and fuse spectral features at different scales to achieve fusion of multi-source features at different scales. In addition, a multi-scale attention decoder is proposed to fuse multi-level road features and global context information. The proposed method was applied to the SpaceNet dataset and self-annotated images from Chongzhou, a representative city in China. Our method performs better over the baseline HRNet by a large margin of +6.38 IoU and +5.11 F1-score on the SpaceNet dataset, +3.61 IoU and +2.32 F1-score on the self-annotated dataset (ChongZhou dataset).

In this article, the exponential synchronization of Markovian jump neural networks (MJNNs) with time-varying delays is investigated via stochastic sampling and looped-functional (LF) approach. For simplicity, it is assumed that there exist two sampling periods, which satisfies the Bernoulli distribution. To model the synchronization error system, two random variables that, respectively, describe the location of the input delays and the sampling periods are introduced. In order to reduce the conservativeness, a time-dependent looped-functional (TDLF) is designed, which takes full advantage of the available information of the sampling pattern. The Gronwall-Bellman inequalities and the discrete-time Lyapunov stability theory are utilized jointly to analyze the mean-square exponential stability of the error system. A less conservative exponential synchronization criterion is derived, based on which a mode-independent stochastic sampled-data controller (SSDC) is designed. Finally, the effectiveness of the proposed control strategy is demonstrated by a numerical example.

Ensemble learning has been used to solve classification and regression problems in many fields, but its application in remote sensing is rarely explored. In this study, an ensemble machine learning algorithm named stacked generalization (or stacking) was developed to forecast surface soil moisture over the Tibetan plateau. The algorithm combined multiple learning techniques including Random Forests (RF), Extreme Gradient Boosting (XGBoost) and linear regression (LR). Normalized difference vegetation index (NDVI), modified soil-adjusted vegetation index (MSAVI), shortwave angle normalized index (SANI) and Shortwave Infrared Transformed Reflectance (STR) derived from Landsat-8 were used as the input variables for the soil moisture retrieving model. The experimental results show that the stacking algorithm is able to achieve an acceptable accuracy, the average root mean squared error (RMSE), mean absolute error (MAE) and coefficient of correlation (r) were 0.052 cm3cm−3, 0.039 cm3cm−3and 0.82 respectively, which is much better than its base models.

Sichuan region as an important hub of western China population, its phenological change has a great influence on the western economic construction and social development, so the phenological response under the background of global warming and change, the ecological balance, scientific research and agricultural production is of great significance [1]. Taking MODIS remote sensing satellite images of Sichuan province as the data set, aiming at the inversion problem of vegetation phenology, the machine learning method--Extreme gradient boosting(XGBoost) was used to build the vegetation phenology prediction model, and the results were compared with the traditional methods. The results show that the prediction model of machine learning method has a certain accuracy. The experimental results show that the XGBoost is able to achieve an acceptable accuracy, the average root mean squared error (RMSE), mean absolute error (MAE) and coefficient of correlation (R) were 4.684/4.413, 4.353/4.297, and 0.7725/0.7812 respectively.

CNN (Convolutional Neural Networks) are inspired by the structure of the visual system and are one of the representative algorithms of deep learning. In recent years, Unet has become an acquaintance of Kaggle Challenge for its simplicity, efficiency and ability to extract features from small_scale samples. Since the model is designed for two classifications, there are serious problems of overfitting in using it for multiple classifications. Specifically, the model fits well in the train set, but there are many missed judgments, false judgments, and speckle noise in the test set. And the categories detected in the result graph are not balanced, some categories have better detection results, and some categories are hardly detected. To solve these problems, this paper proposes a new network that improves Unet: (1) Introduce dropout in the feature extraction stage to prevent overfitting; (2) Introduce ACNet to enhanced feature extraction capabilities. The performance of the new network which was trained with our own training dataset after data augmentation was evaluated with FWIoU and accuracy. Experimental results show that the network has higher FIWoU and accuracy rate under the same data.

Semantic segmentation in high resolution aerial images is a challenging task in remote sensing fields. Compared with other scenarios, semantic segmentation of remote sensing images requires larger receptive fields and more global information. The attention mechanism is one of the most effective way to integrate local features. In this paper, a Triple Attention Network (TANet) is proposed to get more global features. In specific, the paper introduce two self-attention module to get position attention and channel attention. And a label attention module, which generated the attention probability map by introducing spatial context information in the label. The experimental results shows that the proposed network has a higher FWIoU and PA scores than other networks.

Leaf area index (LAI) is very important for crop growth monitoring, biomass estimation, and many plant growth simulation models. Direct methods are the most accurate, but they have the disadvantage of being extremely time-consuming and labor-intensive as a consequence making them not be the preferred solutions. Indirect methods mainly include the radiometric method, inclined point quadrats, and hemispherical canopy photography. This paper focuses on the analysis of the principles and characteristics of the three indirect methods and affirms the advantages and development value of hemispheric canopy photography. The probability model is used to verify the feasibility of LAI inversion which is based on hemispheric canopy photography. In this paper, the canopy images of four sample plots on the campus were collected for LAI estimation, and the results were compared with LAI-2200C. Finally, the improvement of hemispherical canopy photography has prospected.

With the development of deep learning technology, high-resolution remote sensing image change detection based on deep learning has become a hot topic in the field of remote sensing. However, the existing change detection methods based on deep learning only detect the change area of a specific object, and there is no public multi-objects change detection dataset. Focus on these problems, this paper proposed an end-to-end method to obtain the change detection results with change types for high resolution remote sensing images, including sample generation and a deep-learning network, called Res-Unet. Firstly, we obtain the label data by manual annotating. Then, co-registered image pairs are concatenated as an input for the network, and the multi-objects change detection results are directly generated by the network. The experimental results show that the method is effective and Res-Unet has a higher FWloU scores than U-Net.

Land Surface Temperature (LST) is an indicator of the thermal condition at the ground-atmosphere interface. The MODIS LST products from Terra and Aqua satellites provide spatially continuous monitoring ground temperature. But it still has limitations at local scale due to atmospheric disturbances and the spatiotemporal heterogeneity of land surface. In this study, a novel algorithm based on Random Forest (RF) is proposed to reconstruct MODIS LST. The RF was trained using MOD11A1, MOD09A1 and MOD15A2 from Terra MODIS and digital elevation data from ASTER DEM as inputs. LST observations from The Tibetan Plateau Soil Moisture and Temperature Monitoring Network (TP-SMTMN) are used as target data. Experimental results indicate the algorithm can improve the estimation MODIS LST products from the aspects of accuracy and data availability.

DlinkNet[1] (LinkNet With Pretrained Encoder and Dilated Convolution) has been proved to be an effective method for road extraction in remote sensing fields as it won the champion in the DeepGlobe's Road Extraction Challenge. However, as the number of hyperparameters increases (such as the number of channels, filter size, etc.), the difficulty and computational overhead of network design will increase. Focused on this problem, this paper put forward effective ideas to improve D-LinkNet: (1) Applying ResNeXt as its backbone instead of ResNet to rebuild D-LinkNet; (2) Replacing initial block with stem block in the beginning of the network. The results of road extraction which was trained with our own dataset was evaluated with IoU scores. The evaluation results shows that the improved network has higher IoU scores than D-LinkNet when maintaining the model complexity and number of parameters.

The paper described multi-temporal measurements of wheat using ground-based radar scatterometer (GBRS) system and investigated S-band scattering character of wheat parameters to ranks distribution during an entire growth cycle. S-band was chosen for its distinct behavior from popular L and C bands. The difference of backscatter for ranks distribution (row and column) were analyzed as the function of the wheat temporal variations parameters, such as wheat age, biomass, canopy height, LAI at six experimental acquisitions. The research showed the S-band scattering character of wheat parameters to rank distribution has different formulation mechanism and influence on total backscatter. Moreover, the relation between backscatter and ground data had also been analyzed. The research results could be helpful to provide a reference for modeling and cereal monitoring by remote sensing technology.

Quickly and accurately obtaining fuel moisture content information is of great significance for diagnosing vegetation growth, improving agricultural irrigation efficiency, guiding agricultural production, monitoring the drought conditions of natural communities, and forecasting forest fires. Used the measured fuel moisture content in the southern California sample points and various vegetation indices extracted from MODIS remote sensing satellite images as the dataset for the fuel moisture content retrieving model. In this study, three machine learning methods--extreme learning machine (ELM), gradient boosting decision tree (GBDT), and extreme gradient boosting (XGBoost) were used for the fuel moisture content retrieving model. The results show that these three methods can achieve better accuracy than the traditional machine learning method support vector machine (SVM). The experimental results show that the XGBoost is able to achieve an acceptable accuracy, the average root mean squared error (RMSE), mean absolute error (MAE) and coefficient of correlation (R) were 0.1552, 0.1243, and 0.7423 respectively, which is much better than the other models.

Due to different imaging mechanisms, the registration of optical and Synthetic Aperture Radar (SAR) image is a very challenging task. Many optical and SAR registration methods have been proposed. But most of them are for low-to-medium resolution images, and less for high-resolution images. Therefore, this paper proposes a high-resolution optical and SAR image registration method using local self-similar descriptor based on edge feature. Firstly, a Gauss-Gamma bi-windows algorithm is used to extract the edge intensity maps of the images respectively. Its function is to eliminate the non-linear gray-scale difference between SAR and optical images, and also to avoid the interference of isolated speckle noise on feature point extraction. Then, local self-similar descriptor is extracted on the edge intensity map, and descriptor matching is performed using Euclidean distance. Finally, the fast sample consensus algorithm is used to eliminate mismatched point pairs. The experimental results can effectively resist speckle noise and radiation differences, and obtain pixel-level registration accuracy.

This paper proposes a semi-empirical microwave model of methane emissions (CH4) based on the biogeochemical processes from rice paddy. By exploiting the mechanism processes of methane production, oxidation and emission, a combined microwave model is developed to predict methane emissions from rice paddy. Simultaneously, the main influencing emissions factors, which concluded soil, underlying water body, vegetation, climate and management, are analyzed in the present model. During the whole growth season, the multi-polarization backscattering coefficients of rice are measured by the ground-based radar scatterometer (GBRS), and relevant parameters are observed in the rice fields. Moreover, experiments of the emissions samples are conducted on conventional static box, and the results are compared to the semi-empirical microwave model and Denitrification-Decomposition (DNDC) simulation model, respectively, which shows the good extend performance is developed from experience model into mechanism model based on active microwave remote sensing data.

Extracting building information from remote sensing (RS) images have always played an important role in civil and military. In recent years, many efficient approaches are proposed to detect building in remote sensing images. CNN (Convolutional Neural Networks) has proven to be an effective way of this problem. In this paper, to learn building features better, we propose a convolutional network based on Unet++ and containing dense connections. Our contributions are as follows:(1)Rebuild Unet++ by applying DenseNet as its backbone; (2)Add 1×1 convolution layers that can be introduced as bottleneck layer before DenseBlock to reduce the number of input feature-maps, so as reduce the parameters and improve computational efficiency. The accuracy of building extraction results from the new network which was trained with our training dataset after data augmentation was evaluated with IoU scores. The experimental results show the proposed network has higher IoU scores than Unet++ with fewer parameters.

Extracting road information from remote sensing images is one of the hot topics in image processing. The extraction result is saved as raster data, which is difficult to spatial information query, so it's necessary to convert it into vector data. Existing raster data vectorization methods are difficult to maintain the shape of roads and the connection between them, and don't take the attribute information into account. Focus on these problems, this paper proposed an algorithm for raster data vectorization. The algorithm obtains road by tracking pixels in the image, and the road is segmented by nodes (endpoints and intersections), which ensures that their connections in vector data are not disrupted. Then match them with corresponding attribute information (material and width) by image masking. Finally, vector data with attribute information is obtained. The experimental results show that the accuracy of the connection between roads is 97.5% in the vector data obtained by this method, and it has a correct rate of 95% for attribute matching.

Leaf area index (LAI) is the basic factor to understand canopy productivity, soil water evaporation, total transpiration loss, and soil temperature. On the basis of analyzing the merits and demerits of various LAI measurement methods, this paper affirms the development prospect and application value of the hemispherical image method. The paper focuses on the inversion theory of LAI and the extraction of canopy porosity. The essence of the hemispherical image method is further elaborated: after the canopy porosity is obtained by image processing, LAI is retrieved based on Lambert-Beer law. In this paper, four tall arbor forests of Chengdu city are selected as research objects to explore the method of obtaining LAI by hemispherical images and compare with LAI-2200C Plant Canopy Analyzer. The results show that LAI measurement based on the hemispherical image is feasible and credible.

This paper addresses the global stabilization of fractional-order memristor-based neural networks (FMNNs) with time delay. The voltage threshold type memristor model is considered, and the FMNNs are represented by fractional-order differential equations with discontinuous right-hand sides. Then, the problem is addressed based on fractional-order differential inclusions and set-valued maps, together with the aid of Lyapunov functions and the comparison principle. Two types of control laws (delayed state feedback control and coupling state feedback control) are designed. Accordingly, two types of stabilization criteria [algebraic form and linear matrix inequality (LMI) form] are established. There are two groups of adjustable parameters included in the delayed state feedback control, which can be selected flexibly to achieve the desired global asymptotic stabilization or global Mittag-Leffler stabilization. Since the existing LMI-based stability analysis techniques for fractional-order systems are not applicable to delayed fractional-order nonlinear systems, a fractional-order differential inequality is established to overcome this difficulty. Based on the coupling state feedback control, some LMI stabilization criteria are developed for the first time with the help of the newly established fractional-order differential inequality. The obtained LMI results provide new insights into the research of delayed fractional-order nonlinear systems. Finally, three numerical examples are presented to illustrate the effectiveness of the proposed theoretical results.

This paper described the soil scattering measurements from drought status (water content 16.56%) to submerged status (99.92%) at L-, S-, C-, and X-bands. The sensitivity of backscatter to soil water content at multiband was analyzed. Furthermore, the sensitivity of electromagnetic wave to soil water content was discussed in semi-flooded area. The research can be helpful to soil moisture retrieval and supply a reference to soil status monitoring on large scales.

Natural vegetation plays an important role in the ecological environment, because it has become an important indicator to evaluate the status of natural ecosystem. PUE (Precipitation Use Efficiency) has been widely used to evaluate ecosystem by utilizing the ratio of NPP (Net Primary Productivity) to precipitation. Focused on the status of Chinese natural vegetation growth, the precipitation data of 2423 meteorological stations in China from 2000 to 2015 were used to obtain the distribution of regional precipitation by spatial interpolation method. Meanwhile, the NPP spatial distribution data was extracted from MODIS remote sensing data from 2000 to 2015. So, the PUE of national vegetation distribution maps were obtained, and the PUE change trend of six typical natural vegetations was analyzed according to the location in China, which could be applied to evaluate the vegetation ecological status in long time series. The research could be helpful to provide references and evaluation criteria for the spatial-temporal response characteristics of ecological environment.

Recently, D-LinkNet has become a popular convolutional neural network for its high IoU scores in road extraction. Actually, D-LinkNet could have higher IoU scores if it uses ResNet that has deeper network as its encoder part, but the scale of the model would be very huge because of the center part design called DBlock in this paper. Focused on the problem, this paper made some improvements: (1) Add a 1 × 1 convolution that can be introduced as bottleneck layer before DBlock to reduce the number of input feature-maps, so as reduce the parameters and improve computational efficiency. Then add another 1 × 1 convolution after DBlock to recover the required output channels; (2) Rebuild the entire network based on ResNet units with new structure instead of the original one. The new network named as D-LinkNetPlus. Except for reducing parameters, this paper also proposed a method named ESIPs to eliminate small independent patches for the accuracy improvement of road extraction. The experimental results show that the D-LinkNetPlus not only has less parameters but also has a higher accuracy than the results gotten by original structure of D-LinkNet. In addition, the road extraction results presented a better visual effect after processing by ESIPs proposed in the paper.

CNN (Convolutional Neural Networks) has been proved to be an effective method for road extraction in remote sensing fields recently. D-LinkNet based on LinkNet adopted consecutive dilation convolution with different expanding rate to enlarge the receptive field without reducing the resolution of the feature-maps so that had an outstanding performance in high resolution satellite imagery road extraction. However, too many parameters make some inadequacies for D-LinkNet, which introduced LinkNet as its backbone with ResNet construction. Focused on this problem, this paper proposed a new network to improve D-LinkNet: (1) Rebuild D-LinkNet by applying DenseNet as its backbone instead of ResNet; (2) Replacing initial block with stem block in the beginning of the network. The accuracy of road extraction results from the new network which was trained with our own training dataset after data augmentation was evaluated with IoU scores. The experimental results show that the proposed network has a higher IoU scores than D-LinkNet with less parameters.

Interferometric synthetic aperture radar (InSAR) has become the primary means to obtain digital elevation models(DEM) on the earth's surface, including several key steps such as removal of flatten effect, filter processing, and phase unwrapping. Due to atmospheric interference, etc., the distribution of azimuth and range to phase quality on SAR images is uneven. Thus, this paper proposes a method based on weighted quality graph to guide phase unwrapping, which considers the difference of the contribution weight of spatial noise to the range and azimuth of the quality graph. The weighting method is used to eliminate the error of the range direction and the azimuth direction, form a new quality graph, and guide the phase unwrapping. In order to solve the problem of slow speed of traditional methods, this paper introduces the method of heapsort to improve the speed of phase unwrapping.

The extraction of road information has always played a quite important role in civil and military. With the gradual maturity of road extraction technology, how to automatically extract road pavement material information has also begun to attract the attention of researchers. Aiming at this problem, this paper proposes a road material analysis method based on image features and Support Vector Machine (SVM). The method extracts the road surface portion of the road based on the road binary image. Then we use the Rerinex algorithm [1] to denoise the image, the HSV (Hue, Saturation, Value) color model, Local Binary Patterns texture [2], and Gray Level Co-occurrence Matrix (GLCM) are used to extract the features of the road surface pixels. After the principal component analysis reducing the dimension, each feature vector is fused. Then we use the support vector machine (SVM) classifier to analyze the material (asphalt concrete road, cement concrete road and bare soil road) information of the road. This method that combines multiple image features is a new application extension of remote sensing image based on road extraction. The results of the classification experiments on remote sensing images confirmed that the method is effective.

We study the problem of master-slave synchronization of two delayed memristive neural networks (MNNs). Different from most previous papers, memristors are regarded as uncertain continuous time-varying parameters, and MNNs are modeled by neural networks (NNs) with continuous time-varying parameters and polytopic uncertainty. Thus, synchronization of two delayed MNNs is converted into synchronization of delayed NNs with uncertain parameter mismatches. Quasi-synchronization criteria are derived by Lyapunov function and inequality technique. It is shown that, given a predetermined error bound, quasi-synchronization of two delayed chaotic MNNs can be achieved provided that the pinning strength is larger than a threshold. In the end, a numerical example is provided to illustrate the effectiveness of the derived results.

This paper is concerned with quasi-synchronization of delayed memristive neural networks (MNNs) with switching jumps mismatches via aperiodically intermittent control. The issue is presented for three reasons: 1) the existing controllers for synchronization may be too complicated and not economical; 2) under the influence of switching jumps mismatches, synchronization of MNNs may fail to achieve; and 3) matrix measure method is less conservative but cannot be applied directly to synchronization of MNNs. To overcome these difficulties, the concept of asynchronously switching time interval is proposed to describe the phenomenon when the drive-response MNNs switch their connection weights asynchronously. Then, aperiodically intermittent control is designed and quasi-synchronization analysis is carried out based on a combined method that compromises the merits of interval matrix method and matrix measure method. A quasi-synchronization criterion, expressed in terms of the mixture of p-norm and matrix measure of the memristive connection weights, is established. Meanwhile, the fundamental reason for the failure of complete synchronization is revealed. Moreover, an explicit expression of the error level is obtained and the design of the controller under a predetermined error level is presented. The obtained results in this paper reduce the conservativeness and provide a novel insight into the research of synchronization of MNNs.

Cereal crops, contrary to natural vegetation, have the different characteristics for their regular planting. Further, the random assumption of the radiative transfer theory is not suitable for cereal canopy. The paper aimed to present a modified scattering model of row wheat at X-band (center frequency 3.2GHz). The modified scattering model considered both the surface scattering of soil and the volume scattering of wheat canopy. In different wheat growth stage, the weights of the two kinds of scattering phenomenon were set up based on an empirical growth model because of their visible area. A series of data including wheat growth parameters and backscatter coefficients, related to the interaction, were collected for the analyses of the model. The research results showed the model could better reflect the scattering phenomenon of regulate planting, which is helpful to agriculture remote sensing fields.

The paper uses the PSO-SVR algorithm to inverse the water quality parameters based on GF-1 remote sensing image in Longquan lake where is located in Chengdu, Sichuan Province. Longquan Lake is a key drinking water source in Chengdu, so its water quality is very critical. Particle swarm optimization (PSO) optimizes the parameters of the support vector regression (SVR) inversion model to establish the new PSO-SVR inversion model, and PSO can effectively improve the efficiency and the accuracy of the SVR inversion model. At the same, the empirical inversion model was established by using the measured hyperspectral data and concentration of water quality parameters. Comparing with SVR inversion model, PSO-SVR inversion model achieves a better result in the application of suspended solids and Chlorophyll concentration inversion.

Semantic segmentation using adversarial networks has been approved to produce the better artificial results in image processing fields. Focused on current Deep Convolutional Neural Networks (DCNNs), since the convolutional kernel size has been fixed in every convolutional operation, the small objects would be ignored with large convolutional kernel size, and the segmentation result of large objects is not continuous with small convolutional kernel size. The paper developed a semantic segmentation model that combined the adversarial networks with multi-scale context aggregation. Further, the model was applied to road segmentation of UAV RS images. The experimental results of this semantic segmentation model with multi-scale context aggregation has a better performance for road segmentation and fit well with the reference standard results. It can improve the road segmentation accuracy obviously in the situation where there are other small regions whose shape or color is similar to road regions in UAV RS images.

The complex terrain, frequent movement of the earth's crust, dense vegetation exists in the western sichuan plateau where has frequent disasters. PSInSAR can extract the target with stable, strong scattering characteristics to monitor deformation effectively. The experimental area is in maoxian, erlang- mountain, chengdu. Paper uses, Pseudo-3D Phase Unwrapping and LAMBDA Method respectively related to GPS integer ambiguity resolution to unwrap phase and extract the surface deformation of the research area, and comprehensively analyze the effects of different surface and different solutions to the results. Experimental results show that the Network Adjustment Method in elevation correction and linear deformation rate has a comparative advantage, can be more applicable to maoxian, mountain complex mountainous area surrounding the transmission channel of deformation monitoring.

This paper uses the GA-SVM inversion model to invert the suspended matter concentration in Longquan Lake. Genetic algorithm (GA) optimizes the parameters of the SVM inversion model to establish the new GA-SVM inversion model, and GA can effectively improve the efficiency and the accuracy of the SVM inversion model. The inversion model was established by using the measured hyperspectral data and suspended matter concentration. Comparing with SVM inversion model, GA-SVM inversion model achieves a better result in the application of suspended solids concentration inversion. Finally we uses GF-1 remote sensing images to retrieve suspended matter concentration in Longquan lake based on GA-SVM inversion model.

The paper aimed to investigate interaction between electromagnetic wave and cereal parameters at column and row directions. S-band (center frequency 3.2GHz) and wheat had been selected as the research targets. During an entire wheat growth cycle, a series of data related to the interaction were collected for the analyses. Backscatter at row and column directions were analyzed as the function of the wheat temporal variations parameters. The research results show the dominated contribution determined the influence of wheat at row and column directions on radar backscatter. Moreover, a simple model has been presented to minimize the influence of column and row directions on radar backscatter. The research can be helpful to modeling and cereal monitoring by microwave remote sensing technology.

Soil Moisture Content (SMC) plays an important role in different environmental studies. In this study, to assess SMC, a method based on reflectance values in red and NIR bands from HJ-1B is introduced. This model divides SNIR-R into three separate regions based on the pixels NDVI values and fits three different regression equations to each region. All the achieved results showed that as NDVI values get larger, the accuracy of the proposed models decreases. Furthermore, in this study, the model was applied for assessment of SMC. It was concluded that satellite estimated SMC is highly correlated with field measured data.

UAVRS (unmanned aerial vehicles remote sensing) is widely used in various fields such as resource exploration, disaster monitoring. The paper designed a geometric distortion correction algorithm of UAV low altitude remote sensing image for emergency disaster, which mainly consisted of the following steps: eliminating rotational error, extracting the control points effectively and correcting geometric distortion for overlapping regions. Based on comprehensive experiments and applications, the algorithm is proved effectively to reduce the UAV image distortion. Furthermore, the algorithm can also improve the processing quality and reduce the processing speed of the image in emergency disaster situations.

UAV image registration is a key in the process of UAV image matching and application. Based on a large number of images, overlapping irregular and lack of ground control point, paper put forward the image registration algorithm based on overlapping area detection. Firstly the algorithm uses the improved distributed search method based on phase correlation method to search overlapping area and get the translation and rotation parameters of the original image, which is very effective for rotating image. Then a new angular point matching method combined with the offset and rotation angle is used, which is based on Harris. The experimental results show that the method can be applied to UAV remote sensing image registration properly, and has the characteristics of efficiency, accuracy and robustness.

This letter investigates the influence of row wheat for azimuthal look angles on radar backscatter at L-, S-, C-, and X-bands. The radar backscatter was collected with full polarization (HH, HV, VH, and VV) and incidence angles (10°-70°) at different wheat phenological stages. Simultaneously, wheat parameters (biomass, canopy height, stem density, leaf inclination, etc.) and soil parameters (moisture and roughness) were measured to explain the influence based on radiative transfer theory. The research results show that, when the contribution of soil scattering dominates in total backscatter, the influence of row wheat on radar backscatter is varied with the electromagnetic wavelength and the visible soil contained in radar footprint area. When volume scattering contributes more in total backscatter, the influence of row wheat on radar backscatter mainly comes from leaf parameters and wheat geometric shape. Moreover, research results also show that azimuthal look angles affect radar backscatter much mainly for variable scattering cross section and leaf parameters. The research is helpful for cereal monitoring, modeling, and cereal parameter inversion from synthetic aperture radar images.

The research made an adaptation for Michigan Microwave Canopy Scattering (MIMICS) model at four bands (L, S, C, X) after taking wheat ears as the first layer. When wheat ears appears in heading stage, they locate on the top of the whole wheat and owes the different dielectric constant and water content, which should be considered the important scattering elements. The scattering character of wheat ears are fully considered in two growth stage (heading stage and ripening stage). In the process of adaptation, the stem and leaves layer was taken as the second layer instead of the trunk layer, which is different from forested environment to cereal condition. A new contribution to backscatter was inserted to compensate the total backscatter. The results after comparing the measured data and the predicted data by adaptation of MIMICS model showed the inclusion of wheat ears as one of the model components is feasible in modeling the wheat backscatter.

The paper studied the model relation of soil moisture and spectral index by using the data field measured spectral reflectance and soil moisture, and analyzed the correlation between the soil moisture and spectral index. Furthermore, based on analyzing the data measured, the research selected the sensitive band and built the optimal inversion model. The soil moisture of the studied area (Maoergai area of Minjiang upriver) was inversed, and then the inversed results were analyzed and evaluated. By analyzing the different inversed results accuracy, the research obtained the optimal method of soil moisture inversion for ecological water information index remote sensing quantitative model.

This thesis propose an algorithm which is based on the unmanned aerial vehicle (UAV) remote sensing images geometrical rectification model with the flight attitude parameters. The algorithm can make the flight attitude information of the UAV remote sensing images for the fast geometrical distortion rectification in the course of lacking the ground control points to carry on. The quality assessment of the geometrical rectification results of the model is performed through experimental comparison and analysis. Geometrical rectification algorithm can not only obtain high resolution remote sensing images with effective rectification immediately but also improve the quality, speed and accuracy of UAV remote sensing data processing and quantified disaster information extraction.

Based on the remote sensing hyper-spectral reflectance measured in the Lake Kun-cheng in May, 2009 and concurrent chlorophyll-a (Chl-a) concentration assayed in the lab, a three-band model(semi-analytical method), was applied to estimate the concentration of chlorophyll-a to monitor the water quality of the lake Kun-cheng located in Chang-shu city, Jiangsu Province. Compared to the estimating results from the band ratio method based on normalized spectrum and the first derivative method, the results show that the estimating results from the three-band model was more exactly. The three-band model can improve the estimation accuracy for Chl-a by analyzing the inherent optical properties of the chlorophyll-a, the total suspended matter, the CDOM and the pure water and optimizing the choice of the three bands.

Eco-water is defined as a transformation of precipitation, which is deposited by vegetation layer, humiliated vegetation layer and soil layer. It plays an important role in the water-cycle system. As potential factors on Eco-water and Eco-water layer are different from one season to another, multi-temporal and multi-type remote sensing data, measured spectrum and the routine observation were applied to construct the indices for Eco-water and its inversion model. The four Eco-water indices, including Vegetation Canopy Interception Content, Vegetation Water Content Index, Soil Moisture Index and Eco-water Storage Index, were calculated. The results show that the RS information model can reflect the real soil moisture. The dissertation brings forward the Eco-water Remote Sensing quantitative study based on vegetation layer. The vegetation-based calculation model for Eco-water with quantitative remote rensing technology, which has been identified in the dissertation, possesses significant science affect and practical value; and it can not only advance the methods of Eco-environment study, but also promote the research on water-resources transformation and water-cycle, also enlarge the domains of remote sensing applications.