Abstract
We propose a new method for detecting ob jects such as bags carried by pedestrians depicted in short video sequences. In common with earlier work [1,2] on the same problem, the method starts by averaging aligned foreground regions of a walking pedestrian to produce a rep- resentation of motion and shape (known as a temporal template ) that has some immunity to noise in foreground segmentations and phase of the walking cycle. Our key novelty is for carried ob jects to be revealed by comparing the temporal templates against view-specific exemplars generated o?ine for unencumbered pedestrians. A likelihood map ob- tained from this match is combined in a Markov random field with a map of prior probabilities for carried ob jects and a spatial continuity as- sumption, from which we obtain a segmentation of carried ob jects using the MAP solution. We have re-implemented the earlier state of the art method [1] and demonstrate a substantial improvement in performance for the new method on the challenging PETS2006 dataset [3]. Although developed for a specific problem, the method could be applied to the de- tection of irregularities in appearance for other categories of ob ject that move in a periodic fashion.