Depth map calibration by stereo and wireless sensor network fusion

A novel approach to depth map calibration by fusing localization data from wireless sensor networks with depth maps obtained through stereopsis is presented. In experiments, "smart" MICA2 wireless sensors from Crossbow Inc., and a Canon PowerShot SD100 digital camera are used. Sensor locations are determined via an acoustic time-of-flight ranging technique, and the uncalibrated depth map is computed using a binocular stereopsis technique. The fusion is performed (a) by fitting a 3-D surface to a set of a priori known co-planar sensor locations, and (b) by computing the depth map calibration model parameters through minimizing the squared distance between the sensor-defined plane and the corresponding depth map measurements. Algorithms for stereopsis, sensor localization, and depth map and sensor location fusion are presented. Calibration results along with error analysis follows. A summary of challenges with respect to automation, computational requirements, and obtained accuracy of depth estimation conclude the paper.