Object-oriented crop classification based on UAV remote sensing imagery

UAV remote sensing images have the advantages of high spatial resolution,fast speed,strong real-time performance,and convenient operation,etc.,and have become a recently developed,vital means of acquiring surface information.It is an important research task for precision agriculture to make full use of the spectrum,texture,color and other characteristic information of crops,especially the spatial arrangement and structure information of features,to explore effective methods for the classification of multiple varieties of crops.In order to explore the applicability of the object-oriented method to achieve accurate classification of UAV high-resolution images,the paper used the object-oriented classification method in ENVI to classify the UAV high-resolution remote sensing image obtained from the orderly structured 28 species of crops in the test field,which mainly includes image segmentation and object classification.The results showed that the plots obtained after classification were continuous and complete,basically in line with the actual situation,and the overall accuracy of crop classification was 91.73%,with Kappa coefficient of 0.87.Compared with the crop planting area based on remote sensing interpretation and field survey,the area error of 17 species of crops in this study was controlled within 15%,which provides a basis for object-oriented crop classification of UAV remote sensing images.

The paleoclimatic environment reconstruction of Lop Nur in NW China in UAV spectroscopy

The change in the ecological environment in the arid core area is a critical issue in the context of global warming.To study the paleoclimate evolution,precise identification of minerals deposited in Asia’s arid hinterland,Lop Nur Salt Lake,NW China was conducted.The hyperspectral data of the salt crust was sampled to identify the species and content of sedimentary minerals,and the multispectral photos were used to reconstruct the salt crust morphology using the unmanned aerial vehicles platform.The SUnSAL(sparse unmixing by variable splitting and augmented Lagrangian)method was employed to inverse the sedimentary mineral components along the shoreline.The heterogeneity of salt and clay minerals in bright and dark ear-shaped strips was evaluated.The paleoclimatic environment associated with salt lake extinction was reconstructed by analyzing paleoclimate records of sediments,spectral reflectance and morphology of the salt crust.Results show that:(1)the variations in the micro-geomorphology of the salt crust are obviously the reason for the formation of bright and dark ear-shaped strips and the differences in the species and relative content of the sedimentary minerals are the microscopic reason.The high ratio of sedimentary salt minerals to clay minerals(RS/C)contributes to the high reflectivity,and the salt crust presents a bright texture.The low RS/C results in the low reflectivity,salt crust presents a dark texture;(2)the bright and dark ear-shaped strips represent warm-arid and cold-humid climates.The shape of the Lop Nur Lake shoreline evolved due to alternating warm-dry and cold-humid paleoclimate changes.

Geometrical feature analysis and disaster assessment of the Xinmo landslide based on remote sensing data

At 5:39 am on June 24, 2017, a landslide occurred in the village of Xinmo in Maoxian County, Aba Tibet and Qiang Autonomous Prefecture(Sichuan Province, Southwest China). On June 25, aerial images were acquired from an unmanned aerial vehicle(UAV), and a digital elevation model(DEM) was processed. Landslide geometrical features were then analyzed. These are the front and rear edge elevation, accumulation area and horizontal sliding distance. Then, the volume and the spatial distribution of the thickness of the deposit were calculated from the difference between the DEM available before the landslide, and the UAV-derived DEM collected after the landslide. Also, the disaster was assessed using high-resolution satellite images acquired before the landslide. These include Quick Bird, Pleiades-1 and GF-2 images with spatial resolutions of 0.65 m, 0.70 m, and 0.80 m, respectively, and the aerial images acquired from the UAV after the landslide with a spatial resolution of 0.1 m. According to the analysis, the area of the landslide was 1.62 km2, and the volume of the landslide was 7.70 ± 1.46 million m3. The average thickness of the landslide accumulation was approximately 8 m. The landslide destroyed a total of 103 buildings. The area of destroyed farmlands was 2.53 ha, and the orchard area was reduced by 28.67 ha. A 2-km section of Songpinggou River was blocked and a 2.1-km section of township road No. 104 was buried. Constrained by the terrain conditions, densely populated and more economically developed areas in the upper reaches of the Minjiang River basin are mainly located in the bottom of the valleys. This is a dangerous area regarding landslide, debris flow and flash flood events Therefore, in mountainous, high-risk disaster areas, it is important to carefully select residential sites to avoid a large number of casualties.

Detection and threshold-adaptive segmentation of farmland residual plastic film images based on CBAM-DBNet

Robust, accurate, and fast monitoring of residual plastic film (RPF) pollution in farmlands has great significance. Based on CBAM-DBNet, this study proposed a threshold-adaptive joint framework for identifying the RPF on farmland surfaces and estimating its coverage rate. UAV imaging was used to gather images of the RPF from several locations with various soil backgrounds. RPFs were manually labeled, and the degree of RPF pollution was defined based on the RPF coverage rate. Combining differentiable binarization network (DBNet) with the convolutional block attention module (CBAM), whose feature extraction module was improved. A dynamic adaptive binarization threshold formula was defined for segmenting the RPF’s approximate binary map. Regarding the RPF image detection branch, the CBAM-DBNet exhibited a precision (P) value of 85.81%, a recall (R) value of 82.69%, and an F1-score (F1) value of 84.22%, which was 1.09 percentage points higher than the DBNet in the comprehensive index F1 value. For the RPF image segmentation branch, using CBAM-DBNet to segment the RPF image combined with an adaptive binarization threshold formula. Subsequently, the mean absolute percentage error (MAPE), root mean square error (RMSE), and mean absolute error (MAE) of the prediction of RPF’s coverage rate were 0.276, 0.366, and 0.605, respectively, outperforming the DBNet and the Iterative Threshold method. This study provides a theoretical reference for the further development of evaluation technology for RPF pollution based on UAV imaging.

A spatio-temporal multi-scale fusion algorithm for pine wood nematode disease tree detection

Pine wood nematode infection is a devastating disease.Unmanned aerial vehicle(UAV)remote sensing enables timely and precise monitoring.However,UAV aerial images are challenged by small target size and complex sur-face backgrounds which hinder their effectiveness in moni-toring.To address these challenges,based on the analysis and optimization of UAV remote sensing images,this study developed a spatio-temporal multi-scale fusion algorithm for disease detection.The multi-head,self-attention mechanism is incorporated to address the issue of excessive features generated by complex surface backgrounds in UAV images.This enables adaptive feature control to suppress redundant information and boost the model’s feature extraction capa-bilities.The SPD-Conv module was introduced to address the problem of loss of small target feature information dur-ing feature extraction,enhancing the preservation of key features.Additionally,the gather-and-distribute mechanism was implemented to augment the model’s multi-scale feature fusion capacity,preventing the loss of local details during fusion and enriching small target feature information.This study established a dataset of pine wood nematode disease in the Huangshan area using DJI(DJ-Innovations)UAVs.The results show that the accuracy of the proposed model with spatio-temporal multi-scale fusion reached 78.5%,6.6%higher than that of the benchmark model.Building upon the timeliness and flexibility of UAV remote sensing,the pro-posed model effectively addressed the challenges of detect-ing small and medium-size targets in complex backgrounds,thereby enhancing the detection efficiency for pine wood nematode disease.This facilitates early preemptive preser-vation of diseased trees,augments the overall monitoring proficiency of pine wood nematode diseases,and supplies technical aid for proficient monitoring.

Multi-temporal monitoring of wheat growth by using images from satellite and unmanned aerial vehicle

Recently near-ground remote sensing using unmanned aerial vehicles(UAV)witnessed wide applications in obtaining field information.In this research,four Rapideye satellite images and eight RGB images acquired from UAV were used from early June to the end of July,2015 covering two experimental winter wheat fields,in order to monitor wheat canopy growth status and analyze the correlation among satellite images based normalized difference vegetation index(NDVI)with UAV’s RGB images based visible-band difference vegetation index(VDVI)and ground variables of the sampled grain protein contents.Firstly,through image interpretation of UAV’s multi-temporal RGB images with fine spatial resolution,the wheat canopy color changes could be intuitively and clearly monitored.Subsequently,by monitoring the changes of satellite images based NDVI as well as VDVI values and UAV’s RGB images based VDVI values,the conclusions were made that these three vegetation indices demonstrated the same and synchronized trend of increasing at the early stage of wheat growth season,reaching up to peak values at the same timing,and starting to decrease since then.The results of the correlation analysis between NDVI of satellite images and sampled grain protein contents show that NDVI has good predicative capability for mapping grain protein content before ripening growth stage around June7,2015,while the reliability of using satellite image based NDVI to predict grain protein contents becomes worse as ripening stage approaches.The regression analysis between UAV’s RGB image based VDVI and satellite image based VDVI as well as NDVI showed good coefficients of determination.It is concluded that it is feasible and practical to temporally complement satellite remote sensing by using UAV’s RGB images based vegetation indices to monitor wheat growth status and to map within-field spatial variations of grain protein contents for small scale farmlands.