Stereo-camera visual odometry for outdoor areas and in dark indoor environments

A major goal in the research and development area of navigation is to improve the accuracy and the integrity of navigational systems. Scientists and researchers in this field are working on increasing the performance of navigational algorithms while the sensors for navigation systems become smaller and cheaper. Nonetheless ring laser gyroscope (RLG) and fiber optic gyroscope (FOG) are navigational components of the high price segment. For ultra-low-cost navigation systems RLGs and FOGs are too expensive at the present time, so that predominantly MEMS IMUs (microelectromechanical system inertial measurement units) are integrated into such systems. Due to the strong drift of MEMS sensors, aiding sensors have to be used to guarantee the integrity of the system and to get a longterm stable navigation solution. To limit inertial navigation error in outdoor scenarios the usage of Global Navigation Satellite System (GNSS) is an opportunity to deal with this problem. However, it is obvious that the use of GNSS is constrained by the limited availability of satellite signals in urban canyons or indoor situations. Especially close to buildings, GNSS signals are not always available and the information of the satellite signal can be negatively impacted due to multipath and shading. This is why optical sensors like laser range finders or cameras are frequently used as aiding sensors in urban areas or inside buildings.