JPH06187427A - Customer position detecting system - Google Patents

Abstract

PURPOSE:To surely perform the large sale/poor sale management, price analysis, and the commodity management like commodity arrangement by detecting the difference picture between the background picture in a preliminarily set detection area and the active picture at intervals of a certain time and adding it to the preceding difference picture and outputting the calculation result. CONSTITUTION:The video signal from a television camera 3 installed in a salesroom 1 of each floor is successively taken into a customer position detector 5. The background of the detection area of each camera is detected without visiting customers, and next, a sample picture is taken in, and the number of picture elements in each detection area is detected, and the detection result is added to the preceding detection result. When a measurement unit time comes, a communication program between a center personal computer 10 and a local personal computer 11 is started to successively transfer detection data of the local personal computer 11 to the center personal computer 10. Analysis and evaluation are performed by allowing the average number of retained picture elements in each detection area detected in this manner to correspond with POS (point-of-sale) data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to point-of-sale information management (hereinafter referred to as "P" in stores such as department stores and supermarkets).
The present invention relates to a sales floor situation detection system for accurately managing sales / dead sales management, price analysis, and product management such as shelving allocation in association with an OS system.

[0002]

2. Description of the Related Art A POS system installed in a department store supermarket, which is a system related to the present invention, will be described. The POS system is a system that reads the product barcode with a barcode scanner at the POS terminal cashier, transfers the data to the host computer via the telephone line, performs data aggregation processing, and creates a book table. In addition to the code, the store code,
Gondola arrangement of shelves, selling price at store, purchase price, etc., and aggregation processing is performed according to the number of sales of each product. The information processed by the POS system is used for marketing, product planning, sales planning, sales promotion, etc.

Next, a system for detecting the number of customers visiting a store, which is another system related to the present invention, will be described. As for the detection method of this system, the applicant of the present invention applies for Japanese Patent Application No. 3-1640.
In addition to the image processing method presented in Japanese Patent No. 76, an infrared light emitting and receiving element switching method (for example, JP-A-55-16278).
No.), both of which automatically count the number of people entering and leaving the store at the entrances and exits of each floor of the store, such as department stores and supermarkets.
It is a system that calculates the total number of people entering and leaving the store, the staying time, and the average staying time based on the number of people entering and leaving the store per unit time at each entrance and exit, and outputs it as analysis data. From the data of the customer detection system and the POS system, it is possible to find out how many of the customers actually purchased (purchase rate).

[0004]

However, the sales volume data for each product of the POS system is only data at the time of sale and does not take into consideration the situation of the customers visiting the sales floor. Therefore, certain decisions such as marketing, product planning, sales planning, and sales promotion cannot be made. On the other hand, the purchase rate can be obtained by linking the POS system and the store traffic detection system, but the information that one really wants to know cannot be obtained in managing the number of store customers for which product. There was a problem.

In view of the above situation, the present invention provides a device for detecting the position of a customer visiting a store such as a department store or a supermarket, which is effective as product management information, and analyzes it together with the POS system data and the system for detecting the number of customers visiting the store.
It is an object of the present invention to enable reliable decision making by using objective data of the sales floor, as opposed to conventional decision making methods such as price decision, shelving decision, etc. based on the marketing know-how of buyers in department stores and supermarkets.

[0006]

[Means for Solving the Problems] A means for solving the above-mentioned problems is to calculate a TV camera installed in a department store, a supermarket, etc., or an analog video signal from the TV camera into a digital signal for calculation. It is composed of a customer position detection device for processing, detects the difference pixel number between the background image and the current image in a preset detection area at regular time intervals, and adds it with the previously detected difference pixel number in the same detection area. The calculation result is output at a preset time.

[0007]

With the above-mentioned means, it is possible to detect a visitor at a sales floor such as a department store or a supermarket or an exhibition hall by the difference pixel position and the difference pixel number in the detection area per hour.

[0008]

Embodiments of the present invention will be described with reference to the drawings. FIG. 1 shows an embodiment in which the system of the present invention is installed in a department store. First, the configuration will be described. Data communication is performed from a television camera 3 mounted on the ceiling of the sales floor 1 and capturing a detection area around the gondola 2 through a coaxial cable 4, and each floor (in the figure, 1FL and 2FL are taken as an example. Save the detection data (shown in the figure), and connect to the local line 8 through the modem 7.
Center PC 1 installed in the store office 9 using
A local personal computer 11 for data communication with 0 is installed on each floor. In the store office 9, a modem 12 for transferring the detection data of each floor and the detection data are totaled,
A center personal computer 10 for data analysis, a display 13 for displaying and printing analysis results, and a printer 14 are installed.

Next, the operation will be described. The video signal from the TV camera 3 installed in the sales floor 1 on each floor is
With the configuration shown in FIG. 2, the customer position detection devices 5 are sequentially loaded. (From FIG. 2 onward, the same elements as those in FIG. 1 are indicated by the same numbers) Each channel 15a, 15b of the multiplexer 15
The coaxial cables 4a and 4b from the cameras 3a and 3b are connected to each other, and the video signal 15c of the channel designated by the image capture command signal 16a output from the image capture command unit 16 is supplied to the A / D converter 17. It is transferred, converted into a digital signal 17 a, and stored in the image frame memory 18.

Next, a method of performing arithmetic processing based on the image data in the image frame memory 18 will be described. FIG. 3 shows a flow chart of the customer position detection system, which will be described in the order of the flow chart. The detection area is set around the gondola 2 of the captured image shown in FIG. 4 according to the product arrangement as shown by the broken line in the figure. As shown in FIG. 4, the background image of the detection area for each camera is captured in the state where there are no customers, and then the sample image shown in FIG. 5 is captured. 1 in the figure
9 is a visitor. FIG. 6 shows an image that is binarized and compressed by calculating the difference between the background image and the sample image. Reference numeral 20 in the drawing denotes a compressed pixel. From the processed image of FIG. 6, the number of pixels is detected for each detection area, and the detection result is added to the previous detection result. If the above processing is completed for one camera, the processing for the next camera is performed in the same manner, and if the processing for all cameras is completed, the next sampling time is awaited, and similarly, from the first camera signal in sequence. To detect.

When the processing is expressed by a mathematical expression, Here, Ga1 is the average number of staying pixels in the detection area a1, Δ
T is a sampling time (for example, 1/6 second), Gn is the number of detected pixels, and Tn is a detection unit time (for example, 1 hour).

The average number of staying pixels per measurement unit time detected by the above processing is stored in the memory of the local personal computer.

Next, a method of communicating the data stored in the memory of the local personal computer 11 to the center personal computer 10 will be described. FIG. 7 shows a flowchart of data communication. When the measurement unit time comes, the center PC 10
The communication program between the local personal computer 11 and the local personal computer 11 is activated, and the detection data of the local personal computer 11 is sequentially transferred to the center personal computer 11. The average number of staying pixels in each detection area detected as described above and the POS data are made to correspond to each other for analysis / evaluation.

Next, as another embodiment of the present invention, a system in which the system for detecting the number of customers visiting a store and the customer position detecting system of the present invention are combined will be described with reference to the drawings. FIG. 8 shows FIG.
Similarly, with the system installed in the department store floor, the configuration is
The TV cameras 3 are installed in the entrances 21 and 22 of each floor and the sales floor 1, and the video signal from the TV cameras installed in the entrances 21 and 22 are converted and arithmetic processing is performed in the store visitor number detection devices 23 and 24. A customer position detecting device 5 for converting a video signal from the installed TV camera 3 and performing arithmetic processing;
A link cable 25 for connecting the two devices and performing communication of the detected data, and a result of analysis based on the data detected by each device, data from the customer position detecting device 5 via another link cable 26. A local personal computer 11 that communicates with each other, stores the detected data for each floor, and uses a premises line 8 through a modem 7 to communicate with a center personal computer 10 installed in a store office 9 is installed on each floor. The store office 9 is provided with a modem 12 for transferring the detection data of each floor, a center personal computer 10 that collects the detection data and divides the data, a display 13 and a printer 14 for displaying and printing the analysis results. Install.

Next, the operation will be described. FIG. 9 shows a flowchart of this embodiment until the detection data is written to the local personal computer 11 at the measurement unit time on each floor. The left side of the flow chart is the same as the flow chart of the customer position detecting system shown in FIG. On the right side of the flowchart, the applicant of the present invention (Patent
-164076), the flow of the visitor number detection system will be briefly described. The visitor number detection device 2 connected to each of the cameras 3 installed at the entrances 21 and 22.
Processed according to the flow in 3 and 24, as shown in FIG.
The background image of the detection area 27 is taken in when there are no customers, and then the sample image shown in FIG. 11 is taken in. In the figure, 28 is a visitor passing through the detection area. FIG. 12 shows an image in which the difference between the background image shown in FIG. 10 and the sample image shown in FIG. 11 is calculated, binarized, and compressed. The number of compressed pixels for each detection cell 29 and the position of the compressed pixel of each band 30 are detected from the processed image of FIG. 12, and the maximum number of pixels of each cell and each cell are determined from the number of pixels and the position obtained in the previous compressed image. The moving direction of the compressed pixel of is detected and the number of people entering and leaving the store is calculated.
The system for detecting the number of customers visiting the store may be a method other than that of this embodiment as long as it detects the number of people entering and leaving the store.

By the above processing, the measurement time shown in FIG.
3 is obtained, and this data is transferred to the customer position detecting device 5 via the link cable 23.

In the customer position detecting device 5, from the detection data transferred from the store visitor number detecting devices 23 and 24 and the data detected by the customer position detecting device 5, the number of stayers and the average staying amount for each detection area shown in FIG. Calculate time.

When the processing is expressed by calculation, Here, Na1 is the number of stayers in the detection area a1, N is the number of stayers in the entire floor, Ga1 is the average staying pixel number in the detection area a1, and GT is the average staying pixel number in the entire detection area (average staying in each detection area. The added value of the number of pixels) is shown.

[0019] Here, Ta1 is the average residence time of the detection area a1, and T is the average residence time of the entire floor.

The detection data obtained by the above processing is
The data is stored in the memory of the local personal computer 11, and the data is communicated to the center personal computer in the same manner as the data communication explained in FIG. 7 of the above embodiment.

When the detection data collected in the center personal computer are summarized, Is.

[0022]

The effect of the present invention is that the needs of customers who visit the department store supermarket can be surely grasped. The details will be described below. When a customer comes to a department store, supermarket, etc. and purchases a product, the customer with a need visits the leaflet, advertisement, etc., goes to the sales floor of the desired product, and satisfies various selection criteria. It leads to purchase. Here, considering the utilization of the conventional POS system and the detection data obtained by combining the POS system and the customer traffic detection system,
The OS system is based on the data of the sales quantity, sales amount, and the number of customers who purchase each product, and from the viewpoint of grasping customer needs, the data of the POS system is insufficient because it only targets the person who purchased it. , On the other hand, POS
When combining the system and the customer traffic detection system,
By knowing the number of customers visiting the store against the number of customers visiting it, the purchase rate for the entire sales floor can be obtained, but it is not possible to accurately determine whether there is a customer need for each product.

By using the system of the present invention in contrast to this conventional system, it is possible to find out how many stayers exist at each product section, so that the purchase rate for each product can be obtained, and how long the customer stays. Since it is possible to determine whether or not there is a customer need for each product, it is possible to identify the cause of the large or small sales volume (product lineup, price, shelving allocation, etc.). In order to describe the effects arranged in the above more concretely, utilization of data of the system of the present invention will be described in the following items.

(1) Determination of selling and dying muscles Conventional ABC analysis based on POS system data (Amount of sales or gross profit of individual items are arranged in descending order, and products occupying up to 70% of the cumulative ratio are A, 71 to 90%. In the same manner, the data of the number of stayers for each product by the system of the present invention is similarly subjected to ABC analysis, and the judgment result is shown in the table below. Shown in. In the table, when the POS system data and the analysis result by the system data of the present invention match, it is sure that the determination is made only by the conventional POS system data, but when the analysis result does not match, for example, the number of stayers is rank C. If the sales volume is rank A, then increase the number of customers (advertisement /
If the number of stayers is rank A and the number of sales is rank C, there is a customer need, but there is a problem with the sales floor product such as price and quality. If these are adapted to customer needs (price change / product lineup change), the conventional dead-end product can be made into a hot-selling product.

(2) Special sale determination In the conventional POS system, the relationship between price and sales quantity was analyzed and judged, but by looking at the number of people staying in the special sale section, the relationship between the presence or absence of customer needs and the price can be determined. It becomes clear and the special sale can be decided.

(3) Sales decision by time zone By looking at the data of the number of stayers by time zone, the situation of the sales floor can be grasped more quickly in time than the POS system data which is an analysis after purchase. For daily and fresh products, you can make effective sales decisions by time of day.

(4) Price determination In addition to determining the optimal price based on the number of units sold by selling price in the conventional POS system, by dividing the data of the number of stayers together, the price is the criterion for the customer to make a purchase decision at the sales floor. Whether or not can be clarified.

(5) Determination of Shelving Allocation The occupancy data and the average residence time data clarify the customer flow line, and a clear understanding of the customer needs through various analyzes enables a reliable shelving allocation to be determined.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an exemplary embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an image capturing process of a customer position detecting device.

FIG. 3 is a flowchart showing processing of the system of the present invention.

FIG. 4 is an explanatory diagram showing a detection area.

FIG. 5 is an explanatory diagram showing a sample image.

FIG. 6 is an explanatory diagram showing a binarized and compressed image.

FIG. 7 is a flowchart showing a data communication process.

FIG. 8 is a block diagram showing the configuration of another embodiment of the present invention.

FIG. 9 is a flowchart showing the processing of another embodiment.

FIG. 10 is an explanatory diagram showing a background image.

FIG. 11 is an explanatory diagram showing a sample image.

FIG. 12 is an explanatory diagram showing a binarized and compressed image.

FIG. 13 is an explanatory diagram showing detection data.

[Explanation of symbols]

 1 Sales floor 2 Gondola 3 TV camera 5 Customer position detector 7, 12 Modem 8 Private line 9 Store office 10 Center PC 11 Local PC 13 Display 14 Printer

Claims (1)

[Claims] 1. A television camera installed in a department store, a supermarket, etc., or an exhibition hall, and a customer position detecting device for converting an analog video signal from the television camera into a digital signal and processing the digital signal, and at a fixed time interval. ,
The number of difference pixels between the background image and the current image in the detection area set in advance is detected, and the number of difference pixels in the same detection area of the previous time is added, and the calculation result is output at a preset time. A customer position detection system characterized by the above.