Multilevel Structure Extraction-Based Multi-Sensor Data Fusion

Puhong Duan,Xudong Kang,Yu Liu,Pedram Ghamisi

doi:10.3390/rs12244034

Abstract

Multi-sensor data on the same area provide complementary information, which is helpful for improving the discrimination capability of classifiers. In this work, a novel multilevel structure extraction method is proposed to fuse multi-sensor data. This method is comprised of three steps: First, multilevel structure extraction is constructed by cascading morphological profiles and structure features, and is utilized to extract spatial information from multiple original images. Then, a low-rank model is adopted to integrate the extracted spatial information. Finally, a spectral classifier is employed to calculate class probabilities, and a maximum posteriori estimation model is used to decide the final labels. Experiments tested on three datasets including rural and urban scenes validate that the proposed approach can produce promising performance with regard to both subjective and objective qualities.

Highlights

With the advance of imaging techniques, the amount of remote sensing data collected by various remote sensors is growing, which allows us to combine multiple types of data for earth observation [1,2,3]
Method on the light detection and ranging (LiDAR) image can better identify the Building and Water classes compared to hyperspectral image (HSI) (Figure 5)
This is due to the fact that the height and structure information in the LiDAR image make a great contribution in classifying these land covers

Summary

Introduction

With the advance of imaging techniques, the amount of remote sensing data collected by various remote sensors is growing, which allows us to combine multiple types of data for earth observation [1,2,3] Such data record different reflectance characteristics, e.g., rich spectral information, high spatial resolution, and height information. When the land covers are composed of the same material, i.e., roads and roofs, the spectral curves of such objects are very similar. In this situation, it is hard to distinguish these objects with the same spectral curves using HSI data. By integrating the advantages of the two types of data, the fusion of HSI and LiDAR has exhibited better identification performance over single sensor [7,8,9,10]

Objectives

Methods

Results

Discussion

Conclusion