Search

Context information is rarely used in the object-based landcover classification. Previous models that attempted to utilize this information usually required the user to input empirical values for critical model parameters, leading to less optimal performance. Multi-view image information is useful for improving classification accuracy, but the methods to assimilate multi-view information to make it usable for context driven models have not been explored in the literature. Here we propose a novel method to exploit the multi-view information for generating class membership probability. Moreover, we develop a new conditional random field model to integrate multi-view information and context information to further improve landcover classification accuracy. This model does not require the user to manually input parameters because all parameters in the Conditional Random Field (CRF) model are fully learned from the training dataset using the gradient descent approach. Using multi-view data extracted from small Unmanned Aerial Systems (UASs), we experimented with Gaussian Mixed Model (GMM), Random Forest (RF), Support Vector Machine (SVM) and Deep Convolutional Neural Networks (DCNN) classifiers to test model performance. The results showed that our model improved average overall accuracies from 58.3% to 74.7% for the GMM classifier, 75.8% to 87.3% for the RF classifier, 75.0% to 84.4% for the SVM classifier and 80.3% to 86.3% for the DCNN classifier. Although the degree of improvement may depend on the specific classifier respectively, the proposed model can significantly improve classification accuracy irrespective of classifier type. [ABSTRACT FROM AUTHOR]

Traditionally, the multiple images collected by cameras mounted on Unmanned Aircraft Systems (UAS) are mosaicked into a single orthophoto on which Object-Based Image Analysis (OBIA) is conducted. This approach does not take advantage of the Multi-View (MV) information of the individual images. In this study, we introduce a new OBIA approach utilizing multi-view information of original UAS images and compare its performance with that of traditional OBIA, which uses only the orthophoto (Ortho-OBIA). The proposed approach, called multi-view object-based image analysis (MV-OBIA), classifies multi-view object instances on UAS images corresponding to each orthophoto object and utilizes a voting procedure to assign a final label to the orthophoto object. The proposed MV-OBIA is also compared with the classification approaches based on Bidirectional Reflectance Distribution Function (BRDF) simulation. Finally, to reduce the computational burden of multi-view object-based data generation for MV-OBIA and make the proposed approach more operational in practice, this study proposes two window-based implementations of MV-OBIA that utilize a window positioned at the geometric centroid of the object instance, instead of the object instance itself, to extract features. The first window-based MV-OBIA adopts a fixed window size (denoted as FWMV-OBIA), while the second window-based MV-OBIA uses an adaptive window size (denoted as AWMV-OBIA). Our results show that the MV-OBIA substantially improves the overall accuracy compared with Ortho-OBIA, regardless of the features used for classification and types of wetland land covers in our study site. Furthermore, the MV-OBIA also demonstrates a much higher efficiency in utilizing the multi-view information for classification based on its considerably higher overall accuracy compared with BRDF-based methods. Lastly, FWMV-OBIA and AWMV-OBIA both show potential in generating an equal if not higher overall accuracy compared with MV-OBIA at substantially reduced computational costs. [ABSTRACT FROM AUTHOR]

Graphical abstract (A) A general representation of the camera positioning – showing image overlap in red; (B) photos were captured in a parallel-manner to the measured trees within the sampled windbreak and right-of-ways – note the control point target within multiple images; (C) processing the photos created a 3D model of each study site; (D) stem diameter was measured from 'sliced' point cloud data by selecting two points from opposite sides of each stem. Highlights • Tree stems were photographed to create 3D structure from motion (SfM) photogrammetric models. • Diameters at breast height measurements from photogrammetric models were highly correlated with physical measurements. • Photogrammetric point clouds can be used to measure stem DBH, and the point clouds can provide a 3D record the modeled stem. Abstract Much of forest science is dependent on accurate stem measurements, and relatively new photogrammetric techniques may be suitable for modeling stems from the terrestrial perspective. From imagery taken along a windbreak and urban roadways we tested the viability of photogrammetric modeling for producing accurate diameter at breast height measurements. Treatments for different point cloud models differed based on intervals between control points (i.e., every 5 m, 10 m, 25 m, and an absence of target control points) and site conditions (i.e., urban mixed species vs. a windbreak of Pinus taeda) over 100 m sections in the Tampa Bay, FL area. Stem diameter measurements from both the windbreak (n = 53) and the urban sites (n = 93) showed high conformity between field-derived and point cloud model measurements (linear regression showed R2 values >0.9 and RMSE values ranging from 7.04 − 12.35%) with the number of control point targets having little influence on modeled DBH accuracy. Modeled stems of larger trees had greater associated error relative to DBH tape measurements, which can be attributed, in part, to problems with estimating diameter from non-circular stems of certain urban species (i.e., Quercus virginiana). Future work will focus on georeferencing these datasets and extracting data on other aspects of stem biometry (e.g., lean angle of stem, stem volume, etc.). [ABSTRACT FROM AUTHOR]

Florida is the fourth most vulnerable coastal state in the USA to sea level rise (SLR). Studies predict that a 1.20 m rise translates into the displacement of almost five million people and destroys about 2.6 million homes. The only solution to reducing the vulnerability of Florida's coastline is the creation and implementation of coastal policies, including a reduction in armoring and the adoption of policies such as rolling easements. This paper advances a SLR inundation computer model that estimates the costs of applying rolling easement policy through three outcomes: property value loss, property area loss and conservation easement payments to home owners. The GIS computer model is modular which allows new policy components, datasets, or ArcGIS tools to easily be added to the model. The results show that property land inundation and real property losses are primarily linear while rolling easement compensation payments are substantial during the first three scenarios then are largely stable for the remaining SLR steps. [ABSTRACT FROM AUTHOR]

Monitoring aquatic vegetation is an important component of water resource management due to the ecological services provided by these habitats. Spectrally-rich hyperspectral imagery can be an efficient tool for mapping and classifying macrophyte communities. Identification of submerged vegetation in aquatic regions is complicated by variations in optical properties of water constituents, sun-water-sensor geometry, water depth and the spectral/structural complexity of the plants. Many studies have attempted to detect aquatic vegetation in coastal waters; however, few studies have targeted shallow, black-water rivers tainted with chromophoric dissolved organic matter (CDOM). This study investigates methods to analyze airborne hyperspectral imagery and detect and classify aquatic vegetation in a black-water riverine system. Images were normalized to account for reflectance from the water surface and varying water depth before being analyzed by the Maximum Likelihood (ML) and three other non-parametric classifiers: Artificial Neural Network (ANN), Support Vector Machine (SVM) and Spectral Angular Mapper (SAM). Quality assessment analysis indicated a general classification and detection accuracy improvement when non-parametric classifiers were applied on the normalized and depth invariant images. A maximum classification accuracy of about 69% was achieved when the ANN classifier was applied on the normalized images, and maximum detection accuracies of 93% and 92% were obtained when the SAM and the SVM classifiers were applied on depth invariant images, respectively. [ABSTRACT FROM AUTHOR]

Mapping soil nutrients can help smallholder farmers identify soil nutrient status and implement site-specific soil management schemes. In the past, Digital Soil Mapping has seldom been utilized to guide soil nutrient management in smallholder farm settings in South India. The objective of this research was to analyze the spatial resolution effects of different remote sensing images on soil total nitrogen (TN) prediction models in two smallholder villages, Kothapally and Masuti in South India. Regression kriging (RK) was used to characterize the spatial pattern of TN in the topsoil (0–15 cm) by incorporating spectral indices with different spatial resolutions. The results suggested that soil moisture, vegetation, and soil crusts can contribute to the conservation of soil TN in both study areas. Soil prediction models with different spatial resolutions showed a similar spatial pattern of soil TN. The results also demonstrated that the effect of very fine spatial remote sensing spectral data inputs does not always lead to an increase of soil prediction model performance. A RapidEye-based (5 m) soil TN prediction model had lower prediction accuracy than a Landsat 8-based (30 m) soil TN prediction model in Masuti. WorldView-2/GeoEye-1/Pleiades-1A-based (2 m) soil TN prediction models had the highest prediction accuracy in both study areas. The spectral indices based on new bands of WorldView-2 such as coastal, yellow, red edge, and new near infrared bands had relatively strong correlations with soil TN. The utilization of Very High Spatial resolution images such as WorldView-2 in Digital Soil Mapping could improve soil model performance and spatial characterization. Remote sensing-based soil prediction models have high potential to be widely applied in smallholder farm settings. [ABSTRACT FROM AUTHOR]