Majid Banaeyan - Pattern Recognition and Image Processing Group Vienna University of Technology, Vienna, Austria
Faezeh Moteabbed - College of fine arts, University of Tehran

This paper introduces a novel ۳۶۰º calibration tunnel for auto-calibration of multi-camera inthe around-view monitoring system. A specific cylindrical pattern is designed to cover thetunnel and to uniquely encode ۳D real world. The unusual model- and correspondence-freemulti-camera calibration is employed based on our previous work in [۱]. In particular, thescale-free fractal designs inside the ۳۶۰º pattern result in resolution-free calibration and thusmakes the tunnel practical for nearly arbitrary resolutions. One of the main advantage of suchtunnel is to not only avoid of manually place the calibration multiple patterns around a vehiclewhich is complex and laborious but also to dramatically decrease the time of the calibration ina range of few seconds. Therefore, the proposed calibration tunnel can be widely applied tovehicle manufacture, verification, and repair and also in future thanks to the progressing ofartificial intelligence, for installing in every gas station to calibrate the multi-cameras ofautonomous cars.

۳۶۰º Calibration tunnel, Around-view monitoring (AVM), Model- andcorrespondence-free, Multi-camera systems