LaScaBi

Project description

The project "LaScaBi" aims at coregistration of laser scanning data and photogrammetry image or images block and user guidance especially in indoor scene.

Terrestrial Laser Scanning (TLS) can capture highly accurate and dense 3D point clouds from real scenes by sending laser pulses towards targets and measuring the distance from the device to these targets. The ability to quickly and accurately capture the geometry of objects has made TLS a great commercial measurement tool in recent years.  

Despite this, the TLSs also has disadvantages that are evident in many applications:

•    Objects are scanned in a regular pattern (angular steps). Therefore, e.g. the exact position of corners and edges cannot be determined or smaller, complex shaped object sections are not scanned sufficiently dense.
•     Due to the position of the laser scanner a complete coverage of complex objects is not possible, which leads to holes and occlusions in the point cloud.
•     The radiometric quality of TLSs are often not sufficient for a comprehensive semantic description of the objects.

As shown in Figure 1, details of the spiral line are not detected in the laser scan because its resolution is insufficient, the tubes in the upper area are not included in the point cloud because the metal completely reflects the laser beam.

The image, which contains semantic information, can be captured with a high-quality camera from different positions and directions. Therefore, the data captured by laser scanner is supplemented by single images or photogrammetric image mosaics when a more detailed measurement is required or when data in shadowed or hidden areas needs to be captured.

The project aims to develop and test a method with the following features:

• Targeted, selective acquisition of images / image mosaics in areas that cannot be captured by the laser scanner and where image information as such is required.
• Guidance of the user for optimal positioning of the camera.
• Fully automatic registration of the images in the coordinate system realised by the point cloud (immediately after acquisition).

• Direct processing of the hybrid data (images and laser scanner data) in appropriate software.

• Full-fledged functional software