Moving object monitoring surveillance apparatus for detecting, tracking and identifying a moving object by zooming in on a detected flesh color

• US 7,355,627 B2
• Filed: 01/24/2003
• Issued: 04/08/2008
• Est. Priority Date: 01/25/2002
• Status: Active Grant

• Current Assignee: Japan Microsystems Co. Ltd.
• Original Assignee: Japan Microsystems Co. Ltd.
• Inventors: Yamazaki, Sadahiko, Tanibuchi, Kouich
• Primary Examiner(s): Ho; Tuan
• Assistant Examiner(s): Daniels; Anthony J
• Application Number:
  US10/350,832
• Publication Number:
  US 20030142209A1
• Time in Patent Office: 1,901 Days
• Field of Search: 348/143, 348/152, 348/155
• US Class Current: 348/143
• CPC Class Codes: H04N 7/188 Capturing isolated or intermittent images triggered by the occurrence of a predetermined event, e.g. an object reaching a predetermined position signal generation from motion picture films H04N5/253

Abstract

A moving object monitoring surveillance apparatus that controls a camera using a motorized panhead and a motorized zoom lens to identify correctly a moving object, particularly, a person is disclosed. The apparatus includes a video device acquiring information of the image of a moving object as a video signal, a panhead for panning and tilting the video device, a frame grabber for extracting, frame by frame, a still image from the video signal output from the video device and binarizing the still image into frame data, a plurality of frame memories for alternately storing the binarized frame data, a CPU for controlling the operation of the panhead and for processing the frame data, and an image database for storing the processed data. The video device sequentially monitors at least one single surveillance area. The CPU compares frame data with immediately prior frame data to determine a difference therebetween, tracks coordinates corresponding to the difference if the difference is detected, and detects and expands a particular color.

Claims

1. A moving object monitoring surveillance apparatus comprising:

a zoom lens;

a video device acquiring information of the image of a moving object as a video signal;

a panhead for panning and tilting the video device;

a frame grabber for extracting, frame by frame, a still image from the video signal output from the video device and converting the still image into frame data;

a plurality of frame memories for alternately storing the frame data;

a CPU for controlling the operation of the panhead and the zoom lens, and for processing the frame data; and

an image database for storing the processed data, wherein

the CPU retrieves the frame data stored in the frame memory, compares the frame data with immediately prior frame data to detect a difference therebetween, controls the panhead to track coordinates corresponding to the difference when the difference is detected, controls the zoom lens to zoom up a difference image corresponding to the difference on an entire frame, detects a particular color in the difference image zoomed up, and controls the zoom lens to further zoom up the particular color portion of the difference image when the particular color is detected.

2. A moving object monitoring surveillance apparatus according to claim 1, wherein the particular color is a flesh color.

3. A method for zooming in on a particular color portion of a moving object under surveillance, the method comprising the steps of:

a) providing a moving object surveillance apparatus comprising;

a zoom lens;

a video device acquiring information of the image of a moving object as a video signal;

a panhead for panning and tilting the video device;

a frame grabber for extracting, frame by frame, a still image from the video signal output from the video device and converting the still image into frame data;

a plurality of frame memories for alternately storing the frame data;

a CPU for controlling the operation of the panhead and the zoom lens, and for processing the frame data; and

an image database for storing the processed data;

b) monitoring a surveillance area with the moving object surveillance apparatus to detect a moving object within the surveillance area;

c) zooming in on the moving object so that an image of the moving object is expanded within the frame;

d) detecting a particular color in the expanded image of the moving object; and

e) zooming in on a portion of the moving object detected as having the particular color so that an image of the portion of the moving object having the particular color is further expanded within the frame.

4. The method of claim 3 further comprising the step of storing the image of the portion of the moving object having the particular color.

5. The method of claim 3 wherein the moving object is a human.

6. The method of claim 5 wherein the particular color is a flesh color.

7. The method of claim 6 wherein the flesh colored portion of the human is a face.

8. A method for zooming in on an identifying feature of a moving object under surveillance, the method comprising the steps of:

a) providing a moving object surveillance apparatus comprising;

a zoom lens;

a video device acquiring information of the image of a moving object as a video signal;

a panhead for panning and tilting the video device;

a frame grabber for extracting, frame by frame, a still image from the video signal output from the video device and convening the still image into frame data;

a plurality of frame memories for alternately storing the frame data;

a CPU for controlling the operation of the panhead and the zoom lens, and for processing the frame data; and

an image database for storing the processed data; and

b) monitoring a surveillance area with the moving object surveillance apparatus, the step of monitoring including a first whole body tracking mode and a second whole body tracking mode in succession;

wherein in the first whole body tracking mode the CPU retrieves the frame data stored in the frame memory, compares the frame data with the immediately prior frame data to detect a difference therebetween that corresponds to the moving of an object within the surveillance area, controls the zoom lens to zoom in on the moving object so that an image of the moving object is expanded within the frame, and controls the panhead to locate the moving object in the center of the frame and tracks the movement of the moving object thereafter; and

wherein in the second whole body tracking mode the CPU retrieves the frame data of the expanded image of the moving object from the first whole body tracking mode stored in the frame memory, compares the frame data from the first whole body tracking mode with the immediately prior frame data therein to detect a difference in a predetermined particular color corresponding to an identifying feature of the moving object, controls the panhead to locate the identifying feature in the center of the frame, and controls the zoom lens to zoom in on the identifying feature so that an image of the identifying feature is further expanded within the frame.

9. The method of claim 8 wherein the moving object is a human.

10. The method of claim 9 wherein the identifying feature is a face.

11. The method of claim 9 wherein the particular color is a flesh color.

12. The method of claim 10 further comprising the step of storing the frame having the fully expanded image of the identifying feature.