Automatic tracking camera control system

• US 5,434,617 A
• Filed: 12/07/1994
• Issued: 07/18/1995
• Est. Priority Date: 01/29/1993
• Status: Expired due to Term

• Current Assignee: Sunvision Scientific Incorporated
• Original Assignee: Bell Communications Research, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: Bianchi, Michael H.
• Primary Examiner(s): Chin, Tommy P.
• Assistant Examiner(s): LEE, RICHARD J
• Application Number:
  US08/350,752
• Time in Patent Office: 223 Days
• Field of Search: 348/143, 348/155, 348/169, 348/170, 348/171, 348/172, 348/211, 348/214
• US Class Current: 348/170
• CPC Class Codes: G01S 3/7865 using correlation of the live video image with a stored image
• CPC Class Codes: G08B 13/19608 Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position
• CPC Class Codes: G08B 13/19643 wherein the cameras play different roles, e.g. different resolution, different camera type, master-slave camera
• CPC Class Codes: H04N 23/61 based on recognised objects
• CPC Class Codes: H04N 23/695 Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects
• CPC Class Codes: H04N 7/15 Conference systems
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources

Abstract

Methodology and circuitry for automatically effecting electronic camera movement to track and display the location of a moving object, such as a person presenting a talk to an audience. A fixed spotting camera is used to capture a field of view, and a moving tracking camera with pan/tilt/zoom/focus functions is driven to the present location of the moving object. Information for driving the tracking camera is obtained with reference to the pixel difference between a current image and a previous image within the field of view. A tracking algorithm computes the information necessary to drive the tracking camera from these pixel differences as well as data relative to the field of view of the spotting camera and the present tracking camera position.

Claims

1. A method for automatically following a person moving within the view of an electronic camera system comprising a spotting camera and a tracking camera for capturing the picture of the moving person for viewing, the method comprising the steps of:

generating and storing, as a previous image, the image of the person captured by the spotting camera during a scan interval,

generating and storing, as a current image, the image of the person captured by the spotting camera during a scan interval,

determining a threshold box within the field of view of the spotting camera to provide a current threshold value representative of background noise conditions,

determining a current search box from the image captured by the spotting camera to locate the moving person,

determining pixel differences within the current search box between the current image of the person and the previous image, said pixel differences being computed with reference to the current threshold value,

determining a bounding box within the current search box from the pixel differences, the size and position of the bounding box being indicative of the movement detected between the current image of the person and the previous image,

generating a tracking frame with reference to the bounding box, the tracking frame being indicative of the location of the moving person within the view of the spotting camera,

smoothing the size and position of the tracking frame depending on the size and position of the bounding box,

determining a camera frame from the tracking frame and driving the tracking camera to the camera frame, and

adjusting the size of the current search box to closely surround the person as long as the person is moving and to be enlarged when there is no motion by the person.

2. The method in accordance with claim 1 wherein said smoothing step includes calculating a confidence measure based on the number of pixel differences that are above a current threshold within the bounding box and said method further comprising the steps of:

repeating the steps of the method starting with storing the current image of the person as the previous image if the confidence measure is above a minimum measure, and

repeating the steps of the method starting with the previously stored image if the confidence measure is below said minimum measure.