Comparison of Classical Methods and Mask R-CNN for Automatic Tree Detection and Mapping Using UAV Imagery

Detecting and mapping individual trees accurately and automatically from remote sensing images is of great significance for precision forest management. Many algorithms, including classical methods and deep learning techniques, have been developed and applied for tree crown detection from remote sensing images. However, few studies have evaluated the accuracy of different individual tree detection (ITD) algorithms and their data and processing requirements. This study explored the accuracy of ITD using local maxima (LM) algorithm, marker-controlled watershed segmentation (MCWS), and Mask Region-based Convolutional Neural Networks (Mask R-CNN) in a young plantation forest with different test images. Manually delineated tree crowns from UAV imagery were used for accuracy assessment of the three methods, followed by an evaluation of the data processing and application requirements for three methods to detect individual trees. Overall, Mask R-CNN can best use the information in multi-band input images for detecting individual trees. The results showed that the Mask R-CNN model with the multi-band combination produced higher accuracy than the model with a single-band image, and the RGB band combination achieved the highest accuracy for ITD (F1 score = 94.68%). Moreover, the Mask R-CNN models with multi-band images are capable of providing higher accuracies for ITD than the LM and MCWS algorithms. The LM algorithm and MCWS algorithm also achieved promising accuracies for ITD when the canopy height model (CHM) was used as the test image (F1 score = 87.86% for LM algorithm, F1 score = 85.92% for MCWS algorithm). The LM and MCWS algorithms are easy to use and lower computer computational requirements, but they are unable to identify tree species and are limited by algorithm parameters, which need to be adjusted for each classification. It is highlighted that the application of deep learning with its end-to-end-learning approach is very efficient and capable of deriving the information from multi-layer images, but an additional training set is needed for model training, robust computer resources are required, and a large number of accurate training samples are necessary. This study provides valuable information for forestry practitioners to select an optimal approach for detecting individual trees.

Synergistic Use of Geospatial Data for Water Body Extraction from Sentinel-1 Images for Operational Flood Monitoring across Southeast Asia Using Deep Neural Networks

Deep learning is a promising method for image classification, including satellite images acquired by various sensors. However, the synergistic use of geospatial data for water body extraction from Sentinel-1 data using deep learning and the applicability of existing deep learning models have not been thoroughly tested for operational flood monitoring. Here, we present a novel water body extraction model based on a deep neural network that exploits Sentinel-1 data and flood-related geospatial datasets. For the model, the U-Net was customised and optimised to utilise Sentinel-1 data and other flood-related geospatial data, including digital elevation model (DEM), Slope, Aspect, Profile Curvature (PC), Topographic Wetness Index (TWI), Terrain Ruggedness Index (TRI), and Buffer for the Southeast Asia region. Testing and validation of the water body extraction model was applied to three Sentinel-1 images for Vietnam, Myanmar, and Bangladesh. By segmenting 384 Sentinel-1 images, model performance and segmentation accuracy for all of the 128 cases that the combination of stacked layers had determined were evaluated following the types of combined input layers. Of the 128 cases, 31 cases showed improvement in Overall Accuracy (OA), and 19 cases showed improvement in both averaged intersection over union (IOU) and F1 score for the three Sentinel-1 images segmented for water body extraction. The averaged OA, IOU, and F1 scores of the ‘Sentinel-1 VV’ band are 95.77, 80.35, and 88.85, respectively, whereas those of ‘band combination VV, Slope, PC, and TRI’ are 96.73, 85.42, and 92.08, showing improvement by exploiting geospatial data. Such improvement was further verified with water body extraction results for the Chindwin river basin, and quantitative analysis of ‘band combination VV, Slope, PC, and TRI’ showed an improvement of the F1 score by 7.68 percent compared to the segmentation output of the ‘Sentinel-1 VV’ band. Through this research, it was demonstrated that the accuracy of deep learning-based water body extraction from Sentinel-1 images can be improved up to 7.68 percent by employing geospatial data. To the best of our knowledge, this is the first work of research that demonstrates the synergistic use of geospatial data in deep learning-based water body extraction over wide areas. It is anticipated that the results of this research could be a valuable reference when deep neural networks are applied for satellite image segmentation for operational flood monitoring and when geospatial layers are employed to improve the accuracy of deep learning-based image segmentation.

Estimation of Heavy Metals in Agricultural Soils Using Vis-NIR Spectroscopy with Fractional-Order Derivative and Generalized Regression Neural Network

With the development of industrialization and urbanization, heavy metal contamination in agricultural soils tends to accumulate rapidly and harm human health. Visible and near-infrared (Vis-NIR) spectroscopy provides the feasibility of fast monitoring of the variation of heavy metals. This study explored the potential of fractional-order derivative (FOD), the optimal band combination algorithm and different mathematical models in estimating soil heavy metals with Vis-NIR spectroscopy. A total of 80 soil samples were collected from an agriculture area in Suzi river basin, Liaoning Province, China. The spectra for mercury (Hg), chromium (Cr), and copper (Cu) of the samples were obtained in the laboratory. For spectral preprocessing, FODs were allowed to vary from 0 to 2 with an increment of 0.2 at each step, and the optimal band combination algorithm was applied to the spectra after FOD. Then, four mathematical models, namely, partial least squares regression (PLSR), adaptive neural fuzzy inference system (ANFIS), random forest (RF) and generalized regression neural network (GRNN), were used to estimate the concentration of Hg, Cr and Cu. Results showed that high-order FOD had an excellent effect in highlighting hidden information and separating minor absorbing peaks, and the optimal band combination algorithm could remove the influence of spectral noise caused by high-order FOD. The incorporation of the optimal band combination algorithm and FOD is able to further mine spectral information. Furthermore, GRNN made an obvious improvement to the estimation accuracy of all studied heavy metals compared to ANFIS, PLSR, and RF. In summary, our results provided more feasibility for the rapid estimation of Hg, Cr, Cu and other heavy metal pollution areas in agricultural soils.

An index based road feature extraction from LANDSAT-8 OLI images

Road feature extraction from the remote sensing images is an arduous task and has a significant role in various applications of urban planning, updating the maps, traffic management, etc. In this paper, a new band combination (B652) to form a road index (RI) from OLI multispectral bands based on the spectral reflectance of asphalt, is presented for road feature extraction. The B652 is converted to road index by normalization. The morphological operators (top-hat or bottom-hat) uses on RI to enhance the roads. To sharpen the edges and for better discrimination of features, shock square filter (SSF), is proposed. Then, an iterative adaptive threshold (IAT) based online search with variational min-max and Markov random fields (MRF) model are used on the SSF image to segment the roads and non-roads. The roads are extracting by using the rules based on the connected component analysis. IAT and MRF model segmentation methods prove the proposed index (RI) able to extract road features productively. The proposed methodology is a combination of saturation based adaptive thresholding and morphology (SATM), and saturation based MRF (SMRF), applied to OLI images of several urban cities of India, producing the satisfactory results. The experimental results with the quantitative analysis presented in the paper.

Analyzing the Effects of Hyperspectral ZhuHai-1 Band Combinations on LAI Estimation Based on the PROSAIL Model

Leaf area index (LAI) is a key biophysical variable to characterize vegetation canopy. Accurate and quantitative LAI estimation is significant for monitoring vegetation growth status. ZhuHai-1 (ZH-1), which is a commercial remote sensing micro-nano satellite, provides a possibility for quantitative detection of vegetation with high spatial and spectral resolution. However, the band characteristics of ZH-1 are closely related to the accuracy of vegetation monitoring. In this study, a simulation dataset containing 32 bands of ZH-1 was generated by using the PROSAIL model, which was used to analyze the performance of 32 bands for LAI estimation by using the hybrid inversion method. Meanwhile, the effect of different band combinations on LAI estimation was discussed based on sensitivity analysis and the correlation between bands. Then, the optimal band combination from ZH-1 hyperspectral satellite data for LAI estimation was obtained. LAI estimation was performed based on the selected optimal band combination of ZH-1 satellite images in Xiantao city, Hubei province, and compared with the Sentinel-2 normalized difference vegetation index (NDVI) values and LAI product. The results demonstrated that the obtained LAI map based on the optimal band combination of ZH-1 was generally consistent with the overall distribution of Sentinel-2 NDVI and the LAI product, but had a moderate correlation with Sentinel-2 LAI (R = 0.60), which may not favorably indicate the validity of indirect validation. However, the method of this study on the analysis of hyperspectral data bands has application potential to provide a reference for selecting appropriate bands of hyperspectral satellite data to estimate LAI and improve the application of hyperspectral data such as ZH-1 in vegetation monitoring.

SEPARATING THE TERRAIN COVER OF IRAQI MARSHES REGION USINQ NEW SATELLITE BAND COMBINATION

A band combination (542) has been adopted and applied as a new method to classify the Iraqi marshes regions which they are located in the southern of Iraq using Landsat-5 TM scene. The results of proposed band combination were compared to the standard band combinations which they are selected to classify scene classes (541, 543 and 742). In addition, the results reveal that the standard band combinations are failed to discriminate between the scene classes that due to the aquatic nature of the scene, which makes the spectral response of the different classes very close, thus, they miss-classify the scene. Furthermore, the green band which was used in the proposed band combination enhanced the spectral response to discrimination between the different land cover classes. It was found that the support vector machine technique that performed to classify the scenes was revealed to be a very good classifier. The contribution of this study is obvious as the resulting outcomes can be capitalized as guidelines to separate the land cover classes in the aquatic nature to an accuracy that has been reached to 98% compared with the scene’s region of interest.