Search

Under the multiscale distribution, due to dramatic aspect ratio variance leading to prominent arbitrary-oriented character of ships in very high-resolution (VHR) optical remote sensing imagery, how to generate accurate oriented bounding box (OBB) becomes a hot research topic for arbitrary-oriented ship detection. Consequently, in this letter, a concise and effective one-stage anchor-free contour modeling detector called CMDet is proposed for accurate arbitrary-oriented ship detection. Different from currently existed methods via carefully decoupling several independently characteristic parameters for OBB modeling and regression, we resolve the OBB modeling by jointly regressing the contour information. Specifically, the contour information is expressed as a series of Fourier transform coefficients, which are generated by setting up the mapping relation of 1-D Fourier contour coefficients and spatial OBB contour. In addition, a new inherent geometry loss is designed to make detector better learn the geometry information in training phase. After that, the proposed CMDet only needs to predict the correct center point of ships and regress the corresponding entire 1-D Fourier contour coefficients to generate accurate OBB for ship detection. Finally, extensive experiments are carried out on two public OBB ship detection datasets (e.g., HRSC2016 and DIOR-ship), and comparison results demonstrate that the proposed CMDet can obtain the competitive result than the state-of-the-art (SOTA) detectors. [ABSTRACT FROM AUTHOR]

Remote sensing image classification (RSIC) is a classical and fundamental task in the intelligent interpretation of remote sensing imagery, which can provide unique labeling information for each acquired remote sensing image. Thanks to the potent global context information extraction ability of the multi-head self-attention (MSA) mechanism, visual transformer (ViT)-based architectures have shown excellent capability in natural scene image classification. However, in order to achieve powerful RSIC performance, it is insufficient to capture global spatial information alone. Specifically, for fine-grained target recognition tasks with high inter-class similarity, discriminative and effective local feature representations are key to correct classification. In addition, due to the lack of inductive biases, the powerful global spatial context representation capability of ViT requires lengthy training procedures and large-scale pre-training data volume. To solve the above problems, a hybrid architecture of convolution neural network (CNN) and ViT is proposed to improve the RSIC ability, called P 2 FEViT, which integrates plug-and-play CNN features with ViT. In this paper, the feature representation capabilities of CNN and ViT applying for RSIC are first analyzed. Second, aiming to integrate the advantages of CNN and ViT, a novel approach embedding CNN features into the ViT architecture is proposed, which can make the model synchronously capture and fuse global context and local multimodal information to further improve the classification capability of ViT. Third, based on the hybrid structure, only a simple cross-entropy loss is employed for model training. The model can also have rapid and comfortable convergence with relatively less training data than the original ViT. Finally, extensive experiments are conducted on the public and challenging remote sensing scene classification dataset of NWPU-RESISC45 (NWPU-R45) and the self-built fine-grained target classification dataset called BIT-AFGR50. The experimental results demonstrate that the proposed P 2 FEViT can effectively improve the feature description capability and obtain outstanding image classification performance, while significantly reducing the high dependence of ViT on large-scale pre-training data volume and accelerating the convergence speed. The code and self-built dataset will be released at our webpages. [ABSTRACT FROM AUTHOR]

Remote sensing image analysis is a basic and practical research hotspot in remote sensing science. Remote sensing images contain abundant ground object information and it can be used in urban planning, agricultural monitoring, ecological services, geological exploration and other aspects. In this paper, we propose a lightweight model combining vgg-16 and u-net network. By combining two convolutional neural networks, we classify scenes of remote sensing images. While ensuring the accuracy of the model, try to reduce the memory of themodel. According to the experimental results of this paper, we have improved the accuracy of the model to 98%. The memory size of the model is 3.4 MB. At the same time, The classification and convergence speed of the model are greatly improved. We simultaneously take the remote sensing scene image of 64 × 64 as input into the designed model. As the accuracy of the model is 97%, it is proved that the model designed in this paper is also suitable for remote sensing images with few target feature points and low accuracy. Therefore, the model has a good application prospect in the classification of remote sensing images with few target feature points and low pixels. [ABSTRACT FROM AUTHOR]

Currently, under supervised learning, a model pre-trained by a large-scale nature scene dataset and then fine-tuned on a few specific task labeling data is the paradigm that has dominated knowledge transfer learning. Unfortunately, due to different categories of imaging data and stiff challenges of data annotation, there is not a large enough and uniform remote sensing dataset to support large-scale pre-training in the remote sensing domain (RSD). Moreover, pre-training models on large-scale nature scene datasets by supervised learning and then directly fine-tuning on diverse downstream tasks seems to be a crude method, which is easily affected by inevitable incorrect labeling, severe domain gaps and task-aware discrepancies. Thus, in this paper, considering the self-supervised pre-training and powerful vision transformer (ViT) architecture, a concise and effective knowledge transfer learning strategy called ConSecutive Pre-Training (CSPT) is proposed based on the idea of not stopping pre-training in natural language processing (NLP), which can gradually bridge the domain gap and transfer large-scale data knowledge to any specific domain (e.g., from nature scene domain to RSD) In addition, the proposed CSPT also can release the huge potential of unlabeled data for task-aware model training. Finally, extensive experiments were carried out on twelve remote sensing datasets involving three types of downstream tasks (e.g., scene classification, object detection and land cover classification) and two types of imaging data (e.g., optical and synthetic aperture radar (SAR)). The results show that by utilizing the proposed CSPT for task-aware model training, almost all downstream tasks in the RSD can outperform the previous knowledge transfer learning strategies based on model pre-training without any expensive manually labeling and even surpass the state-of-the-art (SOTA) performance without any careful network architecture designing. [ABSTRACT FROM AUTHOR]

In order to accurately segment architectural features in highresolution remote sensing images, a semantic segmentation method based on U-net network multi-task learning is proposed. First, a boundary distance map was generated based on the remote sensing image of the ground truth map of the building. The remote sensing image and its truth map were used as the input in the U-net network, followed by the addition of the building ground prediction layer at the end of the U-net network. Based on the ResNet network, a multi-task network with the boundary distance prediction layer was built. Experiments involving the ISPRS aerial remote sensing image building and feature annotation data set show that compared with the full convolutional network combined with the multi-layer perceptron method, the intersection ratio of VGG16 network, VGG16 + boundary prediction, ResNet50 and the method in this paper were increased by 5.15%, 6.946%, 6.41% and 7.86%. The accuracy of the networks was increased to 94.71%, 95.39%, 95.30% and 96.10% respectively, which resulted in high-precision extraction of building features. [ABSTRACT FROM AUTHOR]

Due to the proliferation of precision agriculture, the obstacle of estimating evapotranspiration (ET) and its components from shadow pixels acquired from remote sensing technology should not be neglected. To accurately detect shaded soil and leaf pixels and quantify the implications of shadow pixels on ET inversion, a two-year field-scale observation was carried out in the growing season for a pinot noir vineyard. Based on high-resolution remote sensing sensors covering visible light, thermal infrared, and multispectral light, the supervised classification was applied to detect shadow pixels. Then, we innovatively combined the normalized difference vegetation index with the three-temperature model to quantify the proportion of plant transpiration (T) and soil evaporation (E) in the vineyard ecosystem. Finally, evaluated with the eddy covariance system, we clarified the implications of the shadow pixels on the ET estimation and the spatiotemporal patterns of ET in a vineyard system by considering where shadow pixels were presented. Results indicated that the shadow detection process significantly improved reliable assessment of ET and its components. (1) The shaded soil pixels misled the land cover classification, with the mean canopy cover ignoring shadows 1.68–1.70 times more often than that of shaded area removal; the estimation accuracy of ET can be improved by 4.59–6.82% after considering the effect of shaded soil pixels; and the accuracy can be improved by 0.28–0.89% after multispectral correction. (2) There was a 2 °C canopy temperature discrepancy between sunlit leaves and shaded leaves, meaning that the estimation accuracy of T can be improved by 1.38–7.16% after considering the effect of shaded canopy pixels. (3) Simultaneously, the characteristics showed that there was heterogeneity of ET in the vineyard spatially and that E and T fluxes accounted for 238.05 and 208.79 W·m−2, respectively; the diurnal variation represented a single-peak curve, with a mean of 0.26 mm/h. Our findings provide a better understanding of the influences of shadow pixels on ET estimation using remote sensing techniques. [ABSTRACT FROM AUTHOR]

Vehicle detection in aerial images is an important and challenging task in the field of remote sensing. Recently, deep learning technologies have yielded superior performance for object detection in remote sensing images. However, the detection results of the existing methods are horizontal bounding boxes that ignore vehicle orientations, thereby having limited applicability in scenes with dense vehicles or clutter backgrounds. In this article, we propose a one-stage, anchor-free detection approach to detect arbitrarily oriented vehicles in high-resolution aerial images. The vehicle detection task is transformed into a multitask learning problem by directly predicting high-level vehicle features via a fully convolutional network. That is, a classification subtask is created to look for vehicle central points and three regression subtasks are created to predict vehicle orientations, scales, and offsets of vehicle central points. First, coarse and fine feature maps outputted from different stages of a residual network are concatenated together by a feature pyramid fusion strategy. Upon the concatenated features, four convolutional layers are attached in parallel to predict high-level vehicle features. During training, task uncertainty learned from the training data is used to weight loss function in the multitask learning setting. For inferencing, oriented bounding boxes are generated using the predicted vehicle features, and oriented nonmaximum suppression (NMS) postprocessing is used to reduce redundant results. Experiments on two public aerial image data sets have shown the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Object detection in remote sensing imagery is a critical yet challenging task in the field of computer vision due to the bird’s-eye-view perspective. Although existing object detection approaches in remote sensing imagery have achieved great advances through the utilization of deep features or rotation proposals, but they give insufficient consideration to multilevel semantic information and its propagation for guiding the learning process. Accordingly, in this article, we propose a hierarchical semantic propagation (HSP) framework to boost object detection performance in remote sensing imagery, which is better able to propagate hierarchical semantic information among different components in a unified network. Given a remote sensing image as input, the HSP framework can detect instances of semantic objects belonging to certain categories in an end-to-end way. First, the multiscale representation is captured by a basic feature pyramid network, which can hierarchically combine spatial attention details and the global semantic structure in order to learn more discriminative visual features. Second, the soft-segmentation prediction is used as an auxiliary objective in the intermediate layer of our HSP; its output instance-aware semantic information can be propagated to suppress noisy background information and thereby guide the proposal generation in the region proposal network. By further propagating this hierarchical semantic information into the region of interest module, we can then predict the object category information and the corresponding horizontal and oriented bounding boxes. Comprehensive evaluations on three benchmark data sets demonstrate the superiority of our HSP to the existing state-of-the-art methods for object detection in remote sensing imagery. [ABSTRACT FROM AUTHOR]

Flash memory is featured with salient advantages over conventional hard disks for massive data storage and efficient on-board data processing. A flash translation layer (FTL) is a critical component for flash-based storage devices to handle particular technical constraints of flash. It is desirable to use flash memory for the storage of massive remote sensing images and support on-board remote sensing data processing applications, which typically require high I/O performance and hence call for advanced FTL design and implementations. In this paper, we introduce our efforts in developing a reinforcement learning driven page-level mapping and caching scheme (named Q-FTL) that is adaptive and responsive to ever-changing I/O streams of on-board remote sensing image processing operations. The adaptability and responsiveness are achieved by the separation of large and small I/O requests, an integrated weighting scheme to measure access costs of cached translation pages, and a reinforcement learning driven cache replacement algorithm. We demonstrate the efficiency of the proposed approach using actual I/O traces generated from on-board remote sensing image processing applications. Experimental results show that Q-FTL improves over several current state-of-the-art FTLs by a large margin and even achieves competitive performance close to an idealized pure page mapping FTL in some cases. [ABSTRACT FROM AUTHOR]

The use of flash-based storage devices has become popular in both research and practice due to their intrinsic advantages over traditional magnetic counterparts. Many advanced flash-aware data management techniques have been developed to exploit the full potential of flash storage devices. It is then reasonable to adopt flash storage devices to deliver high I/O performance for on-board remote sensing image processing tasks, which pose particular challenges in data storage and I/O management. In this work, we focus on caching issues and propose a page-mapping cache management approach based on a page-level flash translation layer (FTL) to accommodate various types of workloads presented by different remote sensing image processing operations. The adaptability is achieved by three major techniques: 1) separating mapping cache into hot and cold regions according to access frequency; 2) a probability-based locality-aware cache replacement priority model to balance high hit ratio and low write/erase costs; and 3) adaptive mapping cache replacement algorithms that are seamlessly integrated with 1) and 2) to handle complicated I/O patterns efficiently. Combining these above three techniques, the proposed FTL-based cache management approach is superior to several existing methods, as demonstrated by extensive experiments running on realistic on-board remote sensing processing I/O traces. [ABSTRACT FROM AUTHOR]

In this study, we investigate the problem of multiclass pixel labeling of very high-resolution (VHR) optical remote sensing images. We propose a novel higher order potential function based on nonlocal shared constraints within the framework of a three-level conditional random field (CRF) model. The proposed approach combines classification knowledge discovery from labeled data with unsupervised segmentation cues derived from the cosegmentation of test data. The cosegmentation of unannotated test data incorporates nonlocal constraints, which are encoded in a novel truncated robust consistency potential function. The class labels are then updated iteratively by alternating between estimating semantic segmentations using CRF and integrating cosegmentation-derived labels in higher order potential functions to refine labeling results. We experimentally demonstrate the improved labeling accuracy of our approach compared with state-of-the-art multilevel CRF approaches based on quantitative and qualitative results. We also show that our approach can address the issue of lacking accurately labeled training data. [ABSTRACT FROM PUBLISHER]

We propose incorporating semantic topic information into a hierarchical conditional random fields (CRFs) framework to promote object recognition and retrieval accuracy. Specially, we devise convenient yet effective methods based on multiple segmentations to perform accurate image retrieval tasks for rigid and amorphous man-made objects. Through a robust topic consistency potential (RTCP) modelling approach, we perform accurate multi-class segmentation on high-resolution remote-sensing images. The generated segments can be readily used for object recognition and discovery. We report satisfactory the performance on two sets of high-resolution remote-sensing images that cover a highly populated urban area and a rural area, respectively. Experimental results demonstrate that our approach outperforms the state-of-the-art CRF models, due to its ability to capture inherent semantic information for efficient object recognition and boundary discovery. [ABSTRACT FROM AUTHOR]

We address the registration problem of multisource three-dimensional (3D) human-made buildings with remote sensing images and the earth's surface in the context of virtual globes. Challenges include fast transformation of 3D coordinates with different reference systems as well as the efficient use of original model information for rigorous and accurate model registration. This paper introduces a novel fast and scalable registration approach that can establish correspondences between heterogeneous external 3D city models and images/terrain surfaces of virtual globes in an efficient and accurate manner. The approach utilizes the projected 3D feature information of 3D city models to develop robust coordinate transformation and reliable model registration methods. The proposed approach builds a solid foundation for the fusion of multisource geospatial data in a united virtual globe reference framework. We report experimental results of online registration tasks for up to over 13K buildings in an integrated 3D virtual globe platform, namely, GeoGlobe. [ABSTRACT FROM PUBLISHER]

In recent years, general-purpose graphics processing units (GP-GPUs) have steadily risen in popularity for remote sensing data processing. Interest has been growing in using hybrid GPU/CPU architectures to realize the full potential of computing devices. This paper studies LiDAR data preprocessing, which is a typical data-intensive remote sensing application. It is proposed to develop an online task scheduler for hybrid GPU/CPU systems using reinforcement learning. At the core of the task scheduler is a Q-learning module that can create the optimal task execution path according to rewards accumulated over time. Constraints and preferences are also encapsulated in the scheduler to support automatic online resource scheduling. Quantitative evaluation on a typical LiDAR data set demonstrates the usefulness and potential of this online task scheduling approach for remote sensing applications. [ABSTRACT FROM PUBLISHER]

Remote sensing images contain complex backgrounds and multi-scale objects, which pose a challenging task for scene classification. The performance is highly dependent on the capacity of the scene representation as well as the discriminability of the classifier. Although multiple models possess better properties than a single model on these aspects, the fusion strategy for these models is a key component to maximize the final accuracy. In this paper, we construct a novel dual-model architecture with a grouping-attention-fusion strategy to improve the performance of scene classification. Specifically, the model employs two different convolutional neural networks (CNNs) for feature extraction, where the grouping-attention-fusion strategy is used to fuse the features of the CNNs in a fine and multi-scale manner. In this way, the resultant feature representation of the scene is enhanced. Moreover, to address the issue of similar appearances between different scenes, we develop a loss function which encourages small intra-class diversities and large inter-class distances. Extensive experiments are conducted on four scene classification datasets include the UCM land-use dataset, the WHU-RS19 dataset, the AID dataset, and the OPTIMAL-31 dataset. The experimental results demonstrate the superiority of the proposed method in comparison with the state-of-the-arts. [ABSTRACT FROM AUTHOR]

• Transpiration and canopy cover of winter wheat was estimated using thermal infrared and visible imagery. • The spatial-temporal variation of transpiration rate and canopy cover under different fertilization levels were quantified and determined. • Relationship between transpiration rate and canopy cover under different fertilization levels was discussed. Food security has always been one of the most important goals for government at all levels in China. Although more fertilization is an effective solution, many other problems appeared with the overuse of fertilization, such as water pollution, soil compaction and so on. Thus, in order to provide suitable advice on fertilization, we used the combination of high-resolution thermal and visible images to evaluate the transpiration rate and canopy cover and identify the effects of fertilization on crop growth. Three different fertilization levels of: (A) 900 kg/ha, (B) 675 kg/ha, and (C) 450 kg/ha were designed for winter wheat. Spatial-temporal distribution of transpiration rate and canopy coverage under different growth stages were depicted based on three temperature model and canopy software. Results showed that: (1) the daily variation of canopy temperature (T c) and transpiration rate were both unimodal curve, and both of them reached their daily peak values at 14:00. Moreover, the spatial heterogeneity of land surface temperature (LST) and transpiration rate was quite obvious and common for all growth stages, and this spatial heterogeneity presented variation from pixel to pixel. (2) Fertilization had a significant effect on the growth of winter wheat. With the higher fertilization amount, the growth of winter wheat became more vigorous, the canopy cover and transpiration rate turned higher, and the T c was lower. For example during recovering stage, the canopy cover under treatment A、B and C were 78.7%, 52.5%, and 42.7%, respectively, as well as the peak values of transpiration rate for treatment A, B, and C were 0.45 mm/h, 0.29 mm/h, and 0.12 mm/h, respectively. The transpiration rate had a positive correlation with canopy cover, and the coefficients of determination R2 = 0.78. (3) The advancement in the crop growth led to the increase of canopy cover and transpiration rate for winter wheat under different fertilization levels. However, the difference of transpiration rate and canopy cover under different fertilization levels decreased gradually with crop growth. Our research accurately detected the effects of fertilization on crop growth based on high-resolution thermal and visible images, which will provide feasible assistance for rational fertilization and it is possible to foresee the broad application prospects in the future. [ABSTRACT FROM AUTHOR]

• We evaluated canopy temperature depression, transpiration and canopy greenness in relation to yield of soybean based on remote sensing. • We identified the optimum time and provide a new way for rapid predicting yield. • CTD and T at noon during the reproductive period can be better used for predicting yield. • Each 1 °C increase in CTD at noon would, on average, reduce yield of soybean by 273–304 kg /ha. Canopy temperature depression (CTD = canopy temperature (T c) – air temperature (T a)), transpiration (T), and canopy greenness (CG) have much to do with crop yields, and they have been widely used to estimate crop yields. However, the issues relating to the best measurement time to predict crop yields have seldom been addressed. Hence, the present study was conducted to identify the best measurement time and provide a new way to rapidly predict soybean yield prediction in a semiarid environment. T c was measured during the reproductive stage under different water stress conditions using a handheld infrared thermal imager that allowed rapid acquisition of high-quality thermal and visible images. T was estimated using the three-temperature model (3 T model) based on thermography, and CG was estimated by analyzing visible images. The results indicate that yield is positively correlated with CG and T to a certain extent; however, it is negatively correlated with CTD. CTD and T at noon during the soybean reproductive period, especially at the flowering and podding stage, are effective in predicting soybean seed yield. During this period, each 1 °C increase in CTD at noon will on average reduce the yield of soybean by 273–304 kg/ha, and when the average T reaches about 1.1 mm/h, the yield no longer increases significantly. Moreover, there is a high correlation between CG (measured by SPAD, Soil-Plant Analysis Development) and the soybean yield during the reproductive stage, especially during the podding and pod-filling stage (R2 = 0.79), which indicates that chlorophyll-based analysis could be used to estimate soybean yield. Therefore, CTD, T, and CG measurements based on remote sensing can be used as key traits to predict soybean yield and make appropriate adaptions to water stress conditions in semiarid areas. [ABSTRACT FROM AUTHOR]

• SoybeanTracer achieved accurate extraction and evaluation of soybean canopy coverage. • SoybeanTracer framework was optimized by introducing in-field scene properties. • SCSNet significantly improved the result of soybean canopy coverage extraction. • Rationality of evaluation system in SoybeanTracer was validated on 196 plots. Soybean is a crucial plant-based protein and vegetable oil source for the global population and a significant cereal-oil dual-purpose crop. Extracting soybean canopy coverage, subsequently implementing evaluation in a low-cost, high-throughput and accurate manner, is highly valuable for soybean breeding and yield increase. Although the utilization of RGB remote sensing images for soybean analysis offers distinct advantages in terms of its cost effectiveness, there are still issues to be addressed. Specifically, the in-field scene properties of soybean canopy data have rarely been developed and there is a gap in automatic evaluation of soybean canopy coverage. For purpose of tackling these pending issues, this study proposed SoybeanTracer, a novel framework designed to achieve accurate soybean canopy coverage extraction and enable specialized evaluation with low cost and high throughput. The proposed framework comprised two components: Soybean Canopy Segmentation Network (SCSNet) and an agronomic soybean coverage evaluation system. Specifically, SCSNet employed a feature extraction network integrating the homogeneity within plot regions and soybean distribution characteristics. To further address the segmentation challenge of soybean plants with sparse growth posed by genotypes, a soybean multiscale feature fusion module was designed. As a result, SCSNet achieved a mean intersection over union (mIoU) and mean pixel accuracy (mPA) of 89.39% and 94.42%, respectively, on soybean canopy segmentation. Therefore, accurate and low-cost extraction of soybean canopy coverage was achieved through the use of RGB remote sensing data. Furthermore, the soybean canopy coverage evaluation system proposed in this study encompassed five levels, characterized by wider spacing at the two ends and narrower spacing in the middle. This agronomic evaluation system had been effectively implemented on 196 plots. The variance in the pairwise levels of the coverage evaluation results demonstrated the robustness of the evaluation system. SoybeanTracer could indeed help to extract soybean canopy coverage and provide valuable reference for implementing soybean genotype selection, laying a useful foundation for other crops in similar scientific tasks. The source code and trained models were available at https://github.com/KLoAS-Research/SCSNet-SoybeanTracer. [ABSTRACT FROM AUTHOR]

• A method was proposed to monitor spatiotemporal variations of transpiration during the whole growth period of crops over heterogeneous farmland. • The estimated transpiration was basically consistent with the measured transpiration, with a high R2 (>0.83) and low RMSE (≤0.1 mm/h). • The mean transpiration of full-irrigated, medium-irrigated, and low-irrigated corn was 0.72, 0.63, and 0.59 mm/h, that of well-irrigated and no-irrigated soybean was 0.77 and 0.27 mm/h. • Although T became slightly underestimated as the pixel scale rose from ground to UAV, the scale effect of estimating T using ground and UAV thermal images is acceptable with small difference between T-ground and T-UAV, especially when the crop canopy is well closed (R2>0.8). Accurate monitoring of crop transpiration (T) is essential for saving and managing irrigation water resources. Ground and UAV (Unmanned Aerial Vehicle) thermal infrared (TIR) images have high enough resolution to meet the needs of precision agriculture. At present, they are mostly used to assess crop water stress instead of to monitor T, especially through UAV TIR images. And the pixel scale effect on T estimation based on ground (T-ground) and UAV (T-UAV) images remains less explored. In this study, the diurnal T of regulated deficit-irrigated corn and soybean was estimated using a three-temperature (3T) model paired with ground and UAV TIR remote sensing images. This method was verified with hydrogen-oxygen stable isotopes measurements, and the pixel scale effect on estimating T with it was analyzed by comparing the LST (Land Surface Temperature) and T derived from ground and UAV TIR images. The results show that the method can accurately estimate T with a high coefficient of determination (R2, 0.84–0.88) and a low root mean square error (RMSE, 0.099–0.104 mm/h) and mean absolute percentage error (MAPE, 12.33 %–12.58 %). Contrary to LST, T-ground and T-UAV decreased as the water stress increased, and their peaks appeared around 15:00. During the experimental period, the mean T of full-irrigated, medium-irrigated, and low-irrigated corn was around 0.72, 0.63, and 0.59 mm/h, respectively, and that of well-irrigated and non-irrigated soybean was 0.77 and 0.27 mm/h, respectively. As the pixel scale increased, the T became slightly underestimated, and the scale effect increased with the complexity of the farmland. This happened mainly because the larger pixel scale of UAV images leads to more mixed pixels and greater soil-vegetation interaction, resulting in overestimation of the canopy temperature. But the T-UAV was highly consistent with the T-ground, with an R2, a RMSE, and a MAPE of 0.60–0.89, 0.06–0.14 mm/h, and 7.39 %–15.51 %, respectively. Therefore, the scale effect on estimating T is acceptable, especially when the canopy is more closed. The proposed method is concluded to be feasible for use in estimating T over heterogeneous farmland. [ABSTRACT FROM AUTHOR]

Soil salinization is a serious land degradation issue in Northwest China and significantly threatens food security; thus, a comprehensive understanding of plant response to salinity stress is desirable. The objective was to detect the effect of salinity stress on alfalfa transpiration rate and plant growth under four salinity levels. The thermal infrared remote (TIR) technique was used to acquire the canopy temperature of plants exposed to four salinities (no salinity, S 0 ; slight salinity, S 3 ; moderate salinity, S 5 ; and severe salinity, S 7), while transpiration was evaluated based on the three temperature (3 T) model. Results indicated that the 3 T model and TIR technique successfully simulated transpiration with a high coefficient determination of R2 above 0.8. The general pattern is that canopy temperature values correspond well with salinity values. With the increase of soil salinity for the treatment of S 3 , S 5 , and S 7 , the canopy temperature increased by 0.88 °C, 0.98 °C, and 1.19 °C, respectively. Moreover, the growth of alfalfa was inhibited, resulting in a corresponding decrease in leaf area index by 19.3%, 31.2%, and 54.2%, respectively. Transpiration reduced by 1.2%, 6.7%, and 20.0%, correspondingly, because of stunted growth and less canopy cover. Salinity stress on agricultural crops was extremely negative; the higher the soil salinity was, the lower the canopy cover, so salinity reduced the growth rate and resulted in reduced plants growing due to osmotic effects. Where salt accumulation occurs in a horizontally nonuniform way, transpiration may be presented by spatial heterogeneity; however, there remains limited research on this issue. This research will provide a scientific basis for ecological construction programs. • We assessed the usefulness of thermal infrared remote technique in detecting transpiration. • Spatial heterogeneity of transpiration was addressed in this research. • The effect of salinity stress on alfalfa transpiration rate and plant growth were quantitatively evaluated. [ABSTRACT FROM AUTHOR]