scholarly journals RESEARCH ON REMOTE SENSING IMAGE CLASSIFICATION BASED ON FEATURE LEVEL FUSION

Author(s):  
L. Yuan ◽  
G. Zhu

Remote sensing image classification, as an important direction of remote sensing image processing and application, has been widely studied. However, in the process of existing classification algorithms, there still exists the phenomenon of misclassification and missing points, which leads to the final classification accuracy is not high. In this paper, we selected Sentinel-1A and Landsat8 OLI images as data sources, and propose a classification method based on feature level fusion. Compare three kind of feature level fusion algorithms (i.e., Gram-Schmidt spectral sharpening, Principal Component Analysis transform and Brovey transform), and then select the best fused image for the classification experimental. In the classification process, we choose four kinds of image classification algorithms (i.e. Minimum distance, Mahalanobis distance, Support Vector Machine and ISODATA) to do contrast experiment. We use overall classification precision and Kappa coefficient as the classification accuracy evaluation criteria, and the four classification results of fused image are analysed. The experimental results show that the fusion effect of Gram-Schmidt spectral sharpening is better than other methods. In four kinds of classification algorithms, the fused image has the best applicability to Support Vector Machine classification, the overall classification precision is 94.01 % and the Kappa coefficients is 0.91. The fused image with Sentinel-1A and Landsat8 OLI is not only have more spatial information and spectral texture characteristics, but also enhances the distinguishing features of the images. The proposed method is beneficial to improve the accuracy and stability of remote sensing image classification.

2020 ◽  
Vol 18 (S3) ◽  
pp. 82-94
Author(s):  
Hongchao Li ◽  
Fang Wu

In this paper, a process visualization model for remote sensing image classification algorithms is constructed to analyze the current processing characteristics of process visualization in remote sensing application systems. The usability of the model is verified in a remote sensing application system with a remote sensing image classification algorithm based on support vector machines as an example. Given the characteristics of remote sensing applications that require high visualization process and a large amount of data processing, the basic process of an image classification algorithm for remote sensing applications is summarized by analyzing the basic process of existing image classification algorithms in remote sensing applications, taking into account the characteristics of process visualization. Based on the existing process of remote sensing image classification algorithm, a process visualization model is proposed. The model takes a goal-based process acts as the basic elements of the model, provides visualization functions and interfaces for human-computer interaction through a human-computer interaction selector, and uses a template knowledge base to save processing data and realize the description of customized processes. The model has little impact on the efficiency and accuracy of the support vector machine-based remote sensing image classification algorithm during the process of process visualization and customization. Finally, the application of the model to integrate business processing of earth observation can address the problem of process customization visualization for remote sensing applications to some extent.


2012 ◽  
Vol 127 ◽  
pp. 237-246 ◽  
Author(s):  
Alexis Comber ◽  
Peter Fisher ◽  
Chris Brunsdon ◽  
Abdulhakim Khmag

Sign in / Sign up

Export Citation Format

Share Document