Exploring the potential of combining time of flight and thermal infrared cameras for person detection

摘要

Combining new, low-cost thermal infrared and time-of-flight range sensorsprovides new opportunities. In this position paper we explore the possibilitiesof combining these sensors and using their fused data for person detection. Theproposed calibration approach for this sensor combination differs from thetraditional stereo camera calibration in two fundamental ways. A firstdistinction is that the spectral sensitivity of the two sensors differssignificantly. In fact, there is no sensitivity range overlap at all. A seconddistinction is that their resolution is typically very low, which requiresspecial attention. We assume a situation in which the sensors' relativeposition is known, but their orientation is unknown. In addition, some of thetypical measurement errors are discussed, and methods to compensate for themare proposed. We discuss how the fused data could allow increased accuracy androbustness without the need for complex algorithms requiring large amounts ofcomputational power and training data.