Scene classification of multisource remote sensing data with two-stream densely connected convolutional neural network

2018 ◽  
Author(s):  
Zhen Wan ◽  
Ronghua Yang ◽  
Yangsheng You ◽  
Zhilin Cao ◽  
Xinan Fang
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Convolutional Neural Network ◽  
Remote Sensing Data ◽  
Scene Classification ◽  
Sensing Data

Neural network classification of remote-sensing data

1995 ◽  
Vol 21 (3) ◽  
pp. 377-386 ◽  
Author(s):  
Diane M. Miller ◽  
Edit J. Kaminsky ◽  
Soraya Rana
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Remote Sensing Data ◽  
Sensing Data ◽  
Neural Network Classification

scholarly journals QUALITY EVALUATION OF LAND-COVER CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK

2018 ◽  
Vol XLII-3 ◽  
pp. 257-262 ◽  
Author(s):  
Y. Dang ◽  
J. Zhang ◽  
Y. Zhao ◽  
F. Luo ◽  
W. Ma ◽  
...  
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Land Cover ◽  
Convolutional Neural Network ◽  
Land Surface ◽  
Quality Evaluation ◽  
Remote Sensing Data ◽  
Land Cover Classification ◽  
Sensing Data ◽  
Evaluation Approach

Land-cover classification is one of the most important products of earth observation, which focuses mainly on profiling the physical characters of the land surface with temporal and distribution attributes and contains the information of both natural and man-made coverage elements, such as vegetation, soil, glaciers, rivers, lakes, marsh wetlands and various man-made structures. In recent years, the amount of high-resolution remote sensing data has increased sharply. Accordingly, the volume of land-cover classification products increases, as well as the need to evaluate such frequently updated products that is a big challenge. Conventionally, the automatic quality evaluation of land-cover classification is made through pixel-based classifying algorithms, which lead to a much trickier task and consequently hard to keep peace with the required updating frequency. In this paper, we propose a novel quality evaluation approach for evaluating the land-cover classification by a scene classification method Convolutional Neural Network (CNN) model. By learning from remote sensing data, those randomly generated kernels that serve as filter matrixes evolved to some operators that has similar functions to man-crafted operators, like Sobel operator or Canny operator, and there are other kernels learned by the CNN model that are much more complex and can’t be understood as existing filters. The method using CNN approach as the core algorithm serves quality-evaluation tasks well since it calculates a bunch of outputs which directly represent the image’s membership grade to certain classes. An automatic quality evaluation approach for the land-cover DLG-DOM coupling data (DLG for Digital Line Graphic, DOM for Digital Orthophoto Map) will be introduced in this paper. The CNN model as an robustness method for image evaluation, then brought out the idea of an automatic quality evaluation approach for land-cover classification. Based on this experiment, new ideas of quality evaluation of DLG-DOM coupling land-cover classification or other kinds of labelled remote sensing data can be further studied.

scholarly journals Application of Artificial Intelligence Algorithms to Control the Use of Flood-prone Areas

2021 ◽  
pp. 6-24
Author(s):  
Keyword(s):  
Neural Network ◽  
Artificial Intelligence ◽  
Remote Sensing ◽  
Convolutional Neural Network ◽  
Satellite Images ◽  
Remote Sensing Data ◽  
Large Spatial Scale ◽  
Sensing Data ◽  
Hazardous Area ◽  
Flood Prone Areas

This article examines the possibility of using artificial intelligence tools to analyze the use of territories prone to flooding during floods. A modern system for monitoring the economic use of flood-prone areas should be based on the use of Earth remote sensing data. The analysis of satellite images, being a laborious task, can be automated through the use of specially trained convolutional neural networks of semantic segmentation based on the algorithm proposed in this article. In this work, on the previously identified flooding zones, using remote sensing data, development objects are automatically determined (segmented) for different times and, by combining information at different times, an assessment of the intensity of this construction in the inter-flood period is made. To form a training sample, a survey of several settlements in the Trans-Baikal Territory was carried out using unmanned aerial vehicles. The neural network was configured using the Python language and the PyTorch library. To select the best convolutional neural network configuration, various combinations of architectures and encoder types were tested for performance and accuracy. The best result in terms of speed and accuracy was shown by the U-Net architecture, built using a convolutional neural network with an SE-ResNeXt50 encoder. According to satellite images of high spatial resolution for the Aginskoye village of Trans-Baikal Kray, a development map was drawn in the flood hazardous area in 2013 and 2019. The objects of development in the period between floods were identified. The results of the study can make it possible to consider a number of important factors when planning the rational use of flood-prone areas in order to improve the quality of life in the region. The obtained maps of the development of flood-prone zones of a large spatial scale are planned to be recommended in the work of state authorities in the field of water resources protection and elimination of natural disasters.

Sign in / Sign up

Export Citation Format

Share Document