Welcome to the Sensor Fusion homepage

Sensor fusion is about mining information from a multitude of sensor measurements, may it be a sensor network or a collection of heterogenous sensors. A smartphone is a good example of a device with many heterogenous sensors, from which added sensor fusion software can compute the orientation of the phone, or even the position inside a building. A sensor  fusion system consists of the following components:

  • An application, including a quantity to be estimated.
  • A multitude of sensors.
  • Signal processing algorithms for pre-processing the sensor data. 
  • A model that relates the various sensor signals to the quantity of interest.
  • Sensor fusion algorithms to combine the information from the individual sensors
  • A recipient of the outputted information, which can be a display, a control system or a decision support system.

The picture below illustrates the signal flow.

This site collects various information sources related to the research and education at Linköping University:

  • The textbooks on Signal Processing and Statistical Signal Processing, respectively, including links to student courses and errata.
  • Software repositories for low-level signal processing functions, as described in the Signal Processing book, and a high-level object oriented Matlab toolbox for Signal and Systems, used to produce the examples and figures in the Sensor Fusion book.
  • Information about the Android Sensor Fusion app, and software repositories for the app.
  • An embryo for a series of video lectures, where one introductory lectures is provided below.

Sensor fusion.