Still image queue analysis system and method

• US 7,652,687 B2
• Filed: 04/18/2007
• Issued: 01/26/2010
• Est. Priority Date: 04/18/2006
• Status: Active Grant

• Current Assignee: Kantar Retail, LLC
• Original Assignee: Sorensen Associates LLC (TNS Group Holdings Ltd.)
• Inventors: Sorensen, Herb
• Primary Examiner(s): CZEKAJ, DAVID J
• Application Number:
  US11/737,071
• Publication Number:
  US 20070253595A1
• Time in Patent Office: 1,014 Days
• Field of Search: 348/150
• US Class Current: 348/150
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G07C 11/00 Arrangements, systems or apparatus for checking, e.g. the occurrence of a condition, not provided for elsewhere
• CPC Class Codes: G07C 2011/02 related to amusement parks
• CPC Class Codes: G07C 2011/04 related to queuing systems
• CPC Class Codes: G07C 9/00 Individual registration on entry or exit

Abstract

A system and method for analyzing a queue. The method may comprise defining a queue and evaluating a series of still images of the queue zone at periodic intervals over a predefined period of time. The method may further comprise approximating an estimated average number of people in the queue over the interval during which the still image was taken to be a number of people counted in the queue zone minus one person who is assumed to be engaged in a transaction. The method may further comprise calculating an estimated average number of the people in the queue over the predefined period by averaging estimated average number of people over intervals within the predefined period. The method may further comprise determining a measure of a queue length based on the estimated average number of people in the queue over the predefined period.

Claims

1. A computerized method for analyzing a queue, comprising:

at an image generating device;

capturing a series of still images of a queue zone at periodic intervals;

at an image analysis device;

evaluating the series of still images of the queue zone taken at the periodic intervals over a predefined period of time;

approximating an estimated average number of people in the queue over the interval during which each still image was taken to be a number of people counted in the queue zone minus one person who is assumed to be engaged in a transaction;

calculating an estimated average number of the people in the queue over the predefined period by averaging the estimated average number of people in the queue over the intervals within the predefined period; and

determining a measure of a queue length based on the estimated average number of people in the queue over the predefined period;

wherein the queue zone includes a wait zone and a transaction zone wherein the transaction zone includes a checkout register;

wherein determining the measure of the queue length is further based on an estimated average period wait time of people in the queue over the predefined period wherein the estimated average period wait time is determined based on information from both the still images and transaction data from the checkout register that provides an elapsed checkout time between every two people in the queue;

wherein determining the estimated average period wait time comprises;

for each still image, approximating an estimated average interval wait time for each still image to be a sum of a check time of each person in the queue, wherein the check time is the elapsed checkout time between two consecutive people; and

averaging the estimated average interval wait time within the predefined period to obtain the estimated average period wait time;

wherein the check time prior to each still image is normalized using the following equation;

normalized check time=check overhead+(check rate×

the number of transaction items)

where the check overhead is an intercept of a line obtained from a regression of check time versus total items in transactions which occurred prior to the still image to be captured.

2. The method of claim 1, further comprising:

determining a still image having a largest number of persons counted within the queue zone over the predefined period; and

approximating an estimated maximum queue length for the queue zone over the predefined period to be equal to the largest number of persons counted in the still image.

3. The method of claim 1, wherein the check overhead and the normalized check time are obtained from medians computed from a large collection of pooled transaction data.

4. The method of claim 3, further comprising determining a normalized average interval wait time of the queue by subtracting an excessive lag time from the estimated average interval wait time wherein the excessive lag time is an aggregate of a difference between an actual check time of each person and the normalized check time.

5. The method of claim 1, wherein the periodic interval for the still image capture is approximately in a range of five minutes to ten minutes and the predefined period is about one hour.

6. The method of claim 1, wherein the still images are captured by a camera.

7. The method of claim 1, wherein still images for a plurality of queue zones are captured by a single camera.

8. The method of claim 1, wherein the queue zone is in an environment selected from the group consisting of retail stores, theme parks, theatres, arenas, stadiums, airports, train stations, bus terminals, toll booths, boarder crossings, airport runways, and ports.

9. A queue analysis system in a designated location, comprising:

at least one image generating device configured to capture a series of still images of people in a queue at periodic intervals over a predefined period of time;

a transaction data generating device configured to record a check time of each person in the queue;

an image analysis system communicating with the image generating device and the transaction data generating device through a network, wherein the image analysis system is configured to;

estimate a queue length to be one of an estimated average number of people in the queue over the predefined period and an estimated average period wait time of people in the queue over the predefined period; and

wherein the estimated average number of the people in the queue over the predefined period is calculated as below;

the estimated average number of the people over the predefined period=

Na=(N₁+N₂+ . . . +Nm)/m

where N₁, N₂, Nm are estimated average numbers of people in the queue over the intervals within the predefined period, m is a number of intervals within the predefined period, and the estimated average number of people in the queue over each interval is approximated to be a number of people counted in the still image for the interval minus one person who is assumed to be engaged in a transaction;

wherein the estimated average period wait time over the predefined period is calculated as below;

T_a=(T_a1+T_a2+ . . . +T_am)/m

where T_a is the estimated average period wait time over the predefined period;

T_a1, T_a2, and T_am are estimated average interval wait times over the intervals within the predefined period and the estimated average interval wait time is approximated to be a sum of an individual check time of each person in the still image wherein the individual check time is an elapsed checkout time between two consecutive people; and

wherein a set of check times prior to the capturing of each still image is normalized using the following equation;

normalized check time=check overhead+(check rate×

a number of transaction items)

where the check overhead is an intercept of a line obtained from a regression of the check time versus total items in a transaction which occurred prior to the capturing of the still image.

10. The queue analysis system of claim 9, wherein a normalized estimated average interval wait time of the queue in one still image is determined by subtracting an excessive lag time from the estimated average interval wait time wherein the excessive lag time is an aggregated difference between an actual check time of a person and the normalized check time.

11. The queue analysis system of claim 9, wherein the image generating device is a camera and the designated location is selected from the group consisting of retail stores, theme parks, theatres, arenas, stadiums, airports, train stations, bus terminals, toll booths, boarder crossings, airport runways, and ports.