Automatic Multiscale Image Acquisition from a Steerable Camera

Abstract

A system for automatically acquiring high-resolution images by steering a pan-tilt-zoom camera at targets detected in a fixed camera view is provided. The system uses automatic or manual calibration between multiple cameras. Using automatic calibration, the homography between the cameras in a home position is estimated together with the effects of pan and tilt controls and the expected height of a person in the image. These calibrations are chained together to steer a slave camera. The manual calibration scheme steers a camera to the desired region of interest and calculates the pan, tile and zoom parameters accordingly.

1 . A method for acquiring multi-scale images, the method comprising:
 tracking an object of interest moving through a field of view of a master camera;   calibrating a steerable slave camera to obtain multi-scale images of the object of interest;   steering the slave camera in accordance with the calibration; and   obtaining multi-scale video of the object of interest using the slave camera.   
     
     
         2 . The method of  claim 1 , wherein the master camera and the slave camera comprise a single camera. 
     
     
         3 . The method of  claim 1 , wherein the step of identifying an object further comprises identifying a plurality of objects and the step calibrating a steerable slave camera comprises calibrating a plurality of steerable cameras, the method further comprising using a camera control policy to associate steerable cameras with objects of interest so as to optimize at least one objective function. 
     
     
         4 . The method of  claim 3 , wherein the objective function comprises the imaged size of an object, the orientation of an object, the location of an object, the length of time an object is viewed, the number of times an object is viewed, the chance of viewing an object before that object is no longer visible, the speed of the object, the type of the object or any other measured parameter of the object, the time taken to steer an available camera to view the object and combinations thereof. 
     
     
         5 . The method of  claim 1 , wherein the step of calibrating the steerable slave camera comprises generating at least one of the pan, the tilt and the zoom parameters for the steerable camera. 
     
     
         6 . The method of  claim 1 , wherein the step of calibrating the steerable slave camera comprises manually calibrating the steerable slave camera. 
     
     
         7 . The method of  claim 6 , wherein the step of manually calibrating the steerable slave camera comprises:
 displaying a field of view of the master camera in a graphical user interface;   identifying the objects of interest in the field of view;   defining at least one region of interest in the field of view;   associating at least one steerable slave camera with each region of interest;   generating pan, tilt and zoom parameters for each steerable camera for objects of interest within the associated region of interest.   
     
     
         8 . The method of  claim 7 , further comprising:
 identifying current and predicted regions of interest through which each object of interest tracks;   selecting steerable slave cameras associated with the identified regions of interest; and   steering the selected slave cameras to the associated regions of interest in accordance with the pan, tilt and zoom parameters generated for the objects of interest that track through the regions of interest.   
     
     
         9 . The method of  claim 1 , wherein the step of calibrating the steerable slave camera comprises automatically calibrating the steerable slave camera. 
     
     
         10 . The method of  claim 9 , wherein the step of automatically calibrating the steerable slave camera comprises:
 establishing a homography between the master camera and the slave camera for objects of interest;   obtaining a height mapping between the master camera and the slave camera for objects of interest; and   obtaining control parameters for the steerable camera.   
     
     
         11 . The method of  claim 10 , wherein the step of establishing a homography comprises:
 directing the master camera and the steerable camera to a common position;   generating a tracking data on a common object for both the master camera and the slave camera;   using the tracking data from the master camera and the slave camera to generate the homography.   
     
     
         12 . The method of  claim 11 , further comprising generating a multiple sets of tracking data, generating at least one homography for each set of tracking data and selecting a best fit homography for use with the steerable slave camera. 
     
     
         13 . The method of  claim 10 , wherein the step of obtaining a height mapping between the master camera and the slave camera comprises determining the height proportionality between the master camera and the slave camera for a common object of interest. 
     
     
         14 . The method of  claim 10 , wherein the step of obtaining control parameters for the steerable camera comprises:
 moving the slave camera in accordance with at least one known pattern;   tracking points along the pattern while the slave camera is moving;   estimating the motion of the slave camera using the tracked points;   comparing the estimated motion of the track points to actual point locations from the known pattern; and   deriving the control parameters from the comparison of estimated points to actual points.   
     
     
         15 . The method of  claim 14 , wherein the step of estimating the motion comprises using affine fit or random sample consensus. 
     
     
         16 . A computer readable medium containing a computer executable code that when read by a computer causes the computer to perform a method for acquiring multi-scale images, the method comprising:
 tracking an object of interest moving through a field of view of a master camera;   calibrating a steerable slave camera to obtain multi-scale images of the object of interest;   steering the slave camera in accordance with the calibration; and   obtaining multi-scale video of the object of interest using the slave camera.   
     
     
         17 . The computer readable medium of  claim 1 , wherein the step of identifying an object further comprises identifying a plurality of objects and the step calibrating a steerable slave camera comprises calibrating a plurality of steerable cameras, the method further comprising using a camera control policy to associate steerable cameras with objects of interest so as to optimize at least one objective function. 
     
     
         18 . The computer readable medium of  claim 1 , wherein the step of calibrating the steerable slave camera comprises manually calibrating the steerable slave camera. 
     
     
         19 . The computer readable medium of  claim 1 , wherein the step of calibrating the steerable slave camera comprises automatically calibrating the steerable slave camera. 
     
     
         20 . A system for indexing multi-scale data, the system comprising:
 a database comprising a plurality of database entries, each database entry associated with at least one multi-scale image;   a control mechanism in communication with the database and capable of retrieving database entries; and   a graphical user interface capable of displaying the multi-scale images associated with database entries retrieved by the control mechanism.