Robust Height Estimation of Moving Objects From Uncalibrated Videos

TitleRobust Height Estimation of Moving Objects From Uncalibrated Videos
Publication TypeJournal Articles
Year of Publication2010
AuthorsShao J, Zhou SK, Chellappa R
JournalIEEE Transactions on Image Processing
Pagination2221 - 2232
Date Published2010/08//
ISBN Number1057-7149
Keywordsalgorithms, Biometry, Calibration, EM algorithm, geometric properties, Geometry, Image Enhancement, Image Interpretation, Computer-Assisted, Imaging, Three-Dimensional, least median of squares, least squares approximations, MOTION, motion information, multiframe measurements, Pattern Recognition, Automated, Reproducibility of results, Robbins-Monro stochastic approximation, robust height estimation, Sensitivity and Specificity, Signal Processing, Computer-Assisted, stochastic approximation, Subtraction Technique, tracking data, uncalibrated stationary camera, uncalibrated videos, uncertainty analysis, vanishing point, video metrology, Video Recording, video signal processing

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.