In response to the existing issues of building edge adhesion, low extraction completeness, and precision when using high-resolution remote sensing images for building extraction, this paper proposes an improved neural network that combines residual connection modules and parallel processing of different-level resolution remote sensing images. This network effectively extracts buildings from high-resolution remote sensing images. On the WHU high-resolution remote sensing image dataset, the proposed method achieves optimal accuracy across all metrics. The precision, recall, F1 score, IoU metric, and mIoU metric of this method are 91.29%, 89.66%, 90.48%, 85.59%, and 88.06%, respectively.
The landslide is a geological disaster that caused the second disruptions and losses next to the earthquake, and people around the world suffered a serious threat to their lives and property damage as a result of landslide disaster every year. So carrying on the study of the landslide hazard will create important theoretical meaning and practical value for proposing targeted prevention and treatment measures. Landslides occur more frequently in Guangxi where serious landslide disasters run riot because of its unique karst topography environment and abundant rainfall. The Wanxiu District of Wuzhou, Guangxi was selected as the study area, the landslide hazard zonation evaluation was studied on the basis of RS and GIS technology. The factors that influence landslide occurrence, such as elevation, slope inclination, slope aspect, curvature and distance to streams were derived from the DEM; land use was extracted from the Google Earth image; lithology was digitalized from the geologic map; rainfall information was from the literature. An improved analytic hierarchy process was presented to determine the index weight in this study. The method weakens the uncertainty in the process of comparing the importance of each factor, it need not to do the consistency check, which can also reduce the iteration times enormously, and improve operating speed. Combined with fuzzy comprehensive evaluation method, the landslide hazard mapping of the study area was made according to the maximum membership degree principle. The resulting map can be used to provide some reference values for risk management, land-use planning and urbanization.
The information extracted from the high spatial resolution remote sensing images has become one of the important data sources of the GIS large scale spatial database updating. The realization of the building information monitoring using the high resolution remote sensing, building small scale information extracting and its quality analyzing has become an important precondition for the applying of the high-resolution satellite image information, because of the large amount of regional high spatial resolution satellite image data. In this paper, a clustering segmentation classification evaluation method for the high resolution satellite images of the typical rural buildings is proposed based on the traditional KMeans clustering algorithm. The factors of separability and building density were used for describing image classification characteristics of clustering window. The sensitivity of the factors influenced the clustering result was studied from the perspective of the separability between high image itself target and background spectrum. This study showed that the number of the sample contents is the important influencing factor to the clustering accuracy and performance, the pixel ratio of the objects in images and the separation factor can be used to determine the specific impact of cluster-window subsets on the clustering accuracy, and the count of window target pixels (Nw) does not alone affect clustering accuracy. The result can provide effective research reference for the quality assessment of the segmentation and classification of high spatial resolution remote sensing images.
This paper design and implement security monitor system within a scenic spot for tourists, the scenic spot staff can be automatic real time for visitors to perception and monitoring, and visitors can also know about themselves location in the scenic, real-time and obtain the 3D imaging conditions of scenic area. Through early warning can realize "parent-child relation", preventing the old man and child lost and wandering. Research results to the further development of virtual reality to provide effective security early warning platform of the theoretical basis and practical reference.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.