A large part of the projects in Visual Sweden are based on the availability of imaging sensors for image analysis and visualization. 3D modeling technologies have been successfully demonstrated in several Visual Sweden projects and used to an increasing extent in various applications. 3D models are developed from camera data, laser radar, thermal imaging but also ground penetrating radar. An example of using 3D models from National Forensic Center (NFC) is a high resolution 3D model of Drottninggatan in Stockholm, as shown in Figure 1, which was used in the investigation and trial of the terrorist attack in April 2017. The model of Drottninggatan is based on laser radar (lidar). The model is textured (colored) based on standard photographs and shows the appearance of the environment in visible light. Another technique that can be used is to create 3D models based on a large number of photographs taken from ground or air (photogrammetry).
The purpose of the proposed platform project is to enable collaboration between algorithm developers and end users, such as companies and more applied research projects. This will be done by facilitation of method refinement and adaptation to specific needs, sensor types and applications. In addition it will introduce new sensor types and thus add more dimensions to the models than spatial information and color. This is achieved by combinations of multidimensional sensor informatics and intelligent algorithms into “intelligent multidimensional imaging”. This concept also offers unique access to a laboratory environment that provides both knowledge, sensors and algorithms with the possibility of imaging a scene in more dimensions (n-dimensional imaging). Due to the availability, cost and complexity of implementation, a similar environment can be very difficult to create for an individual organization or company. The goal is to provide the partners in the project with enough knowledge to be able to bring the results to the market within a short time frame. This requires exchange of knowledge, development of algorithms for using the sensor data, introduction of new sensors, and also a good knowledge of market needs. Since no actor in Sweden alone can provide all this knowledge, the project will provide an environment where the competences of the participating partners can be combined. A plan for how this environment can be created will be developed in the beginning of the project, based on the end-user needs.
The platform project will thus further develop today’s high-resolution, precise 3D models into unique and intelligent n-dimensional models for analysis and visualization of complex environments and events and thus provide means for new products and services for the participating industries whilst feeding back ideas for further research topics to academia.
Projektet drivs av
RISE, FOI, NFC & CVL
januari 2019 - december 2020