Object Reconstruction Using TomoSAR Point Clouds

Project Leader
Xiaoxiang Zhu

Contact Person
Muhammad Shahzad

Cooperation Partners
German Aerospace Center (DLR)
Technical University Munich (TUM)

 

From a stack of SAR images, differential SAR tomography retrieves elevation and deformation information (linear, seasonal, etc.) of possible multiple scatterers inside a single SAR image pixel. It is so far the most competent interferometric SAR methods for uran monitoring. Our TomoSAR software: Tomo-GENESIS produces 4-D point cloud of the illuminated area with point density of about 1 million points/km2. Please find more detail in the "4D City" project.
Object reconstruction from these high quality TomoSAR point clouds can greatly support the reconstruction of dynamic city models that could potentially be used to monitor and visualize the dynamics of urban infrastructure in very high level of details. Motivated by this, Figure 1 presented the first result of building façade reconstruction from TomoSAR point cloud. Only a single view (ascending stack) was used in this study, and therefore, only one side of the building façades can be reconstucted.

tomosar vegas

facade vegas

Figure 1. Upper: TomoSAR point cloud of Las Vegas. Color reprents height. (unit: meter) Lower: 3-D view of the final façade reconstruction.

 

Figure 2 shows the complete reconstructed 3-D model of Bellagio Hotel in Las Vegas on which the amplitude of seasonal deformation is projected to show the periodic movement of the whole façade. In this work, both ascending and descending TomoSAR point cloud were used. The basic idea of such reconstruction is to reconstruct 3-D building models via independent modeling of each individual façade to build the overall 2-D shape of the building footprint followed by its representation in 3-D.

object reconstruction tomosar

Figure 2. Reconstructed 4-D building model of Bellagio Hotel in Las Vegas. Projected on the façades are the color coded amplitude of seasonal motion. (unit: millimeters)

Signal Processing in Earth Observation
Prof. Xiaoxiang Zhu

Technische Universität München
Arcisstr. 21
D-80333 München

 

Latest news

 

Upcoming Events

Research Seminar
13:00, 31/05/2017
FRINGE 2017
05/06/2017
ISPRS Hannover Workshop 2017
06/06/2017
CVPR 2017
21/07/2017