Visual Localization and Identification of Vehicles Inside a Parking House.

The paper discusses a method for simultaneous localization and identification of vehicles by active visual ID tags - small LED arrays. Lights from ID tags are captured by the standard camera which allows low cost of the overall solution, high flexibility of changing ID if needed and resistance against changes in ambient light. The proposed localization algorithm is capable of detecting not only position but also direction of the vehicle. Algorithm has been implemented in the OpenCV framework and tested on multiple E-puck mobile robotic platforms. Output of the algorithm can be used for real-time navigation of the vehicle inside parking house, applicable for both indoor and outdoor environments. [ABSTRACT FROM AUTHOR]