TY - JOUR T1 - Robust Height Estimation of Moving Objects From Uncalibrated Videos JF - IEEE Transactions on Image Processing Y1 - 2010 A1 - Jie Shao A1 - Zhou,S. K A1 - Chellapa, Rama KW - algorithms KW - Biometry KW - Calibration KW - EM algorithm KW - geometric properties KW - Geometry KW - Image Enhancement KW - Image Interpretation, Computer-Assisted KW - Imaging, Three-Dimensional KW - least median of squares KW - least squares approximations KW - MOTION KW - motion information KW - multiframe measurements KW - Pattern Recognition, Automated KW - Reproducibility of results KW - Robbins-Monro stochastic approximation KW - robust height estimation KW - Sensitivity and Specificity KW - Signal Processing, Computer-Assisted KW - stochastic approximation KW - Subtraction Technique KW - tracking data KW - uncalibrated stationary camera KW - uncalibrated videos KW - uncertainty analysis KW - vanishing point KW - video metrology KW - Video Recording KW - video signal processing AB - This paper presents an approach for video metrology. From videos acquired by an uncalibrated stationary camera, we first recover the vanishing line and the vertical point of the scene based upon tracking moving objects that primarily lie on a ground plane. Using geometric properties of moving objects, a probabilistic model is constructed for simultaneously grouping trajectories and estimating vanishing points. Then we apply a single view mensuration algorithm to each of the frames to obtain height measurements. We finally fuse the multiframe measurements using the least median of squares (LMedS) as a robust cost function and the Robbins-Monro stochastic approximation (RMSA) technique. This method enables less human supervision, more flexibility and improved robustness. From the uncertainty analysis, we conclude that the method with auto-calibration is robust in practice. Results are shown based upon realistic tracking data from a variety of scenes. VL - 19 SN - 1057-7149 CP - 8 M3 - 10.1109/TIP.2010.2046368 ER -