Attribute-based person tracking across multiple cameras

摘要

Techniques for tracking an individual across two or more cameras are provided. The techniques include detecting an image of one or more individuals in each of two or more cameras, tracking each of the one or more individuals in a field of view in each of the two or more cameras, applying a set of one or more attribute detectors to each of the one or more individuals being tracked by the two or more cameras, and using the set of one or more attribute detectors to match an individual tracked in one of the two or more cameras with an individual tracked in one or more other cameras of the two or more cameras.

主权项

1. A method for tracking an individual across two or more cameras, wherein the method comprises:
detecting an image of one or more individuals in each of two or more cameras; tracking each of the one or more individuals in a field of view in each of the two or more cameras; applying a set of multiple attribute detectors to the images of each of the one or more individuals being tracked by the two or more cameras, wherein said attribute detectors (i) provide a probability that given attributes are present in an image and (ii) are learned from training images in multiple levels of resolution to produce robustness to multiple changes in lighting and viewpoint; and using the set of multiple attribute detectors to match an individual tracked in one of the two or more cameras with an individual tracked in one or more other cameras of the two or more cameras, wherein using the set of multiple attribute detectors to match an individual tracked in one of the two or more cameras with an individual tracked in one or more other cameras of the two or more cameras comprises:
using a maximum confidence value of each of the multiple attribute detectors to generate a feature vector of each of the one or more individuals;determining a weight associated with each of the multiple attribute detectors, wherein said weight corresponds to a reliability measure of the corresponding attribute detector relative to the remaining attribute detectors in the set of multiple attribute detectors;calculating a distance between each vector using a weighted vector distance between each vector based on the determined weight associated with each of the multiple attribute detectors; andcomparing the distance to a threshold to determine if the individual tracked in one of the two or more cameras is the same individual as the individual tracked in one or more other cameras of the two or more cameras.