The conventional method for crop insect detection based on visual judgment of the field is time-consuming, laborious, subjective, and error prone. The early detection and accurate localization of agricultural insect pests can significantly improve the effectiveness of pest control as well as reduce the costs, which has become an urgent demand for crop production. Maize Spodoptera frugiperda is a migratory agricultural pest that has severely decreased the yield of maize, rice, and other kinds of crops worldwide. To monitor the occurrences of maize Spodoptera frugiperda in a timely manner, an end-to-end Spodoptera frugiperda detection model termed the Pest Region-CNN (Pest R-CNN) was proposed based on the Faster Region-CNN (Faster R-CNN) model. Pest R-CNN was carried out according to the feeding traces of maize leaves by Spodoptera frugiperda. The proposed model was trained and validated using high-spatial-resolution red–green–blue (RGB) ortho-images acquired by an unmanned aerial vehicle (UAV). On the basis of the severity of feeding, the degree of Spodoptera frugiperda invasion severity was classified into the four classes of juvenile, minor, moderate, and severe. The degree of severity and specific feed location of S. frugiperda infestation can be determined and depicted in the frame forms using the proposed model. A mean average precision (mAP) of 43.6% was achieved by the proposed model on the test dataset, showing the great potential of deep learning object detection in pest monitoring. Compared with the Faster R-CNN and YOLOv5 model, the detection accuracy of the proposed model increased by 12% and 19%, respectively. Further ablation studies showed the effectives of channel and spatial attention, group convolution, deformable convolution, and the multi-scale aggregation strategy in the aspect of improving the accuracy of detection. The design methods of the object detection architecture could provide reference for other research. This is the first step in applying deep-learning object detection to S. frugiperda feeding trace, enabling the application of high-spatial-resolution RGB images obtained by UAVs to S. frugiperda-infested object detection. The proposed model will be beneficial with respect to S. frugiperda pest stress monitoring to realize precision pest control.

To obtain effective indoor moving target localization, a reliable and stable moving target localization method based on binocular stereo vision is proposed in this paper. A moving target recognition extraction algorithm, which integrates displacement pyramid Horn–Schunck (HS) optical flow, Delaunay triangulation and Otsu threshold segmentation, is presented to separate a moving target from a complex background, called the Otsu Delaunay HS (O-DHS) method. Additionally, a stereo matching algorithm based on deep matching and stereo vision is presented to obtain dense stereo matching points pairs, called stereo deep matching (S-DM). The stereo matching point pairs of the moving target were extracted with the moving target area and stereo deep matching point pairs, then the three dimensional coordinates of the points in the moving target area were reconstructed according to the principle of binocular vision’s parallel structure. Finally, the moving target was located by the centroid method. The experimental results showed that this method can better resist image noise and repeated texture, can effectively detect and separate moving targets, and can match stereo image points in repeated textured areas more accurately and stability. This method can effectively improve the effectiveness, accuracy and robustness of three-dimensional moving target coordinates.

Rapid urban expansion has brought new challenges to firefighting, with the speed of firefighting rescue being crucial for the safety of property and life. Thus, fire prevention and rescuing people in distress have become more challenging for city managers and emergency responders. Unfortunately, existing research does not consider the negative effects of the current spatial distribution of fire-risk areas, land cover, location, and traffic congestion. To address these shortcomings, we use multiple methods (including geographic information system, multi-criterion decision-making, and location–allocation (L-A)) and multi-source geospatial data (including land cover, point-of-interest, drive time, and statistical yearbooks) to identify suitable areas for fire brigades. We propose a method for identifying potential fire-risk areas and to select suitable fire brigade zones. In this method, we first remove exclusion criteria to identify spatially undeveloped zones and use kernel density methods to evaluate the various fire-risk zones. Next, we use analytic hierarchy processes (AHPs) to comprehensively evaluate the undeveloped areas according to the location, orography, and potential fire-risk zones. In addition, based on the multi-time traffic situation, the average traffic speed during rush hour of each road is calculated, a traffic network model is established, and the travel time is calculated. Finally, the L-A model and network analysis are used to map the spatial coverage of the fire brigades, which is optimized by combining various objectives, such as the coverage rate of high-fire-risk zones, the coverage rate of building construction, and the maintenance of a sub-five-minute drive time between the proposed fire brigade and the demand point. The result shows that the top 50% of fire-risk zones in the central part of Wuhan are mainly concentrated to the west of the Yangtze River. Good overall rescue coverage is obtained with existing fire brigades, but the fire brigades in the north, south, southwest, and eastern areas of the study area lack rescue capabilities. The optimized results show that, to cover the high-fire-risk zones and building constructions, nine fire brigades should be added to increase the service coverage rate from 93.28% to 99.01%. The proposed method combines the viewpoint of big data, which provides new ideas and technical methods for the fire brigade site-selection model.

The detection of moving objects in images is a crucial research objective; however, several challenges, such as low accuracy, background fixing or moving, ‘ghost’ issues, and warping, exist in its execution. The majority of approaches operate with a fixed camera. This study proposes a robust feature threshold moving object identification and segmentation method with enhanced optical flow estimation to overcome these challenges. Unlike most optical flow Otsu segmentation for fixed cameras, a background feature threshold segmentation technique based on a combination of the Horn–Schunck (HS) and Lucas–Kanade (LK) optical flow methods is presented in this paper. This approach aims to obtain the segmentation of moving objects. First, the HS and LK optical flows with the image pyramid are integrated to establish the high-precision and anti-interference optical flow estimation equation. Next, the Delaunay triangulation is used to solve the motion occlusion problem. Finally, the proposed robust feature threshold segmentation method is applied to the optical flow field to attract the moving object, which is the. extracted from the Harris feature and the image background affine transformation model. The technique uses morphological image processing to create the final moving target foreground area. Experimental results verified that this method successfully detected and segmented objects with high accuracy when the camera was either fixed or moving.

Accurately identifying and obtaining changes in ecosystem drivers and the spatial heterogeneity of their impacts on ecosystem services can provide comprehensive support information for ecological governance. In this study, we investigate the changes in the relationship between human and natural factors and water-related ecosystem services (WESs) in different sub-watersheds across various time periods, focusing on four aspects: single-factor effect, nonlinear effect, interactive effects, and spatial characteristics. Taking the southern basins, which have complex topographic, climatic, and economic characteristics, as a study area, the study area was divided into four sub-basins with different characteristics. WESs of water yield, soil conservation, and water purification were quantified using the InVEST model for five periods from 2000 to 2020, and the OPGD and MGWR models were integrated to assess the impacts of 15 factors on WESs and their spatial characteristics. The results show the following: (1) After comparing the data over multiple time periods, climate factors such as precipitation (0.4033) are the primary factors affecting WESs in the southern basins, and human factors such as construction area (0.0688) have a weaker influence. The direct impact of human factors on WESs is not significant in the short term but increases over time. (2) Different sub-watersheds have different impacts on WESs. For instance, human activity intensity (0.3518) is a key factor affecting WESs in the Inward Flowing Area, while precipitation is the primary factor influencing WESs in other sub-watersheds. (3) Influencing factors and WES changes are often nonlinearly correlated; however, once a certain threshold is exceeded, they may have adverse impacts on WESs. (4) When a single factor interacts with other factors, its explanatory power tends to increase. (5) Compared to traditional methods, the estimation accuracy of MGWR is higher. Intense human activities can adversely affect WESs, while abundant precipitation creates favorable conditions for the formation of WESs. Therefore, integrating long-time-series multi-remote sensing data with OPGD and MGWR models is suitable for identifying and analyzing the driving mechanisms of human and natural factors that influence changes in WESs. Against the backdrop of global change, elucidating the driving factors of ecosystem services can provide crucial insights for developing practical policies and land management applications.