JPH07262274A - Automatic collecting device for customer data - Google Patents

Abstract

PURPOSE:To automatically collect customer data without a person by sorting customers into prescribed classes based on an extracted feature and a knowledge base and storing them. CONSTITUTION:A video camera 1 is installed on an automatic vending machine so as to color-photograph a customer perchasing a commodity by the automatic vending machine. The image pickup signal of the video camera 1 is converted to digital data by an A/D conversion part 2 and after then, written into a frame memory 3. Then, preprocessed image data is sent to a feature part extracting part 6 so as to extract the feature part of a person. Image data of each feature part is sent to a feature quantity extracting part 7. The feature quantity extracting part 7 extracts a feature quantity for judging the distinction of sex based on image data of each feature part. Then based on each feature quantity, the features are classified and a judging part 8 generally judges the distinction of sex. Information on the estimated distinction of sex, etc., is put in order by a CPU part 12 and recorded by a storage part 10.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a vending machine,
The present invention relates to a customer data automatic collection device that images a customer with a video camera installed at a cash register or the like and automatically collects customer data based on the captured image.

[0002]

2. Description of the Related Art Conventionally, image processing techniques have been widely used for positioning automatic machines and inspecting parts. In addition, the image processing technology is used to increase the efficiency of elevator operation by counting the number of people in the elevator hall by image processing. Further, image processing technology is also used for face identification (see Japanese Patent Publication No. 6-7388).

However, a method utilizing image processing technology for collecting market data such as customer data has not been developed yet.

[0004]

Conventionally, in order to statistically analyze and analyze the purchaser demographic for a product in product planning and the like, by using a human tactic, for example, by waiting near the vending machine all day long. It collects customer data.

It is an object of the present invention to provide a customer data collecting device which can automatically collect customer data by utilizing image processing technology.

[0006]

A first customer data automatic collecting apparatus according to the present invention extracts an image of a customer, performs image processing on the image, and extracts the characteristics of the customer. On the basis of this, there is provided a classifying means for classifying the photographed customers into predetermined classes, and a storage means for storing a classification result by the classifying means.

A second customer data automatic collection device according to the present invention is a noise reduction device for an image pickup device for photographing a customer, an image memory for storing a customer image input from the image pickup device, and a customer image stored in the image memory. Pre-processing means for performing pre-processing such as removal and feature enhancement, extraction means for extracting the customer's characteristics from the pre-processed image, and a predetermined number of customers based on the extracted characteristics and knowledge base. It is characterized in that it is provided with a classifying means for classifying into categories and a storage means for storing a classification result.

[0008]

In the first customer data automatic collection device according to the present invention, the customer is imaged by the imaging device. By subjecting the captured image to image processing, the characteristics of the customer are extracted, and the photographed customers are classified into predetermined categories based on the extracted characteristics. Then, the classification result is stored in the storage means.

In the second customer data automatic collection device according to the present invention, the image of the customer is picked up by the image pickup device. The customer image input from the imaging device is stored in the image memory. Pre-processing such as noise removal and feature enhancement is performed on the customer image stored in the image memory. The feature amount of the customer is extracted from the preprocessed image. Customers are classified into predetermined categories based on the extracted features and the knowledge base. Then, the classification result is stored in the storage means.

[0010]

【Example】

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows an automatic customer data collection device.

The CPU section 12 controls each section via the system bus 21. The image data of each unit is transferred at high speed via the image bus 22.

The video camera 1 is installed in an automatic vending machine, and takes a color image of a customer who purchases a product from the automatic vending machine. The video camera 1 may be always driven, or the video camera 1 may be driven for a predetermined period of time each time the customer selects the product selection button. The image pickup signal of the video camera 1 is converted into digital data by the A / D converter 2 and then written in the frame memory 3.

The image data stored in the frame memory 3 is sent to the preprocessing unit 5 via the image bus 22 and subjected to noise removal, feature enhancement, rotation processing and the like. Pretreatment unit 5
Is output to the feature extraction unit 9 directly via the image bus 22 or written in the image memory 4.

When it is difficult to perform the pre-processing at one time, the frame memory 3, the image bus 22, the pre-processing section 5, the image bus 2
2, the image memory 4, the image bus 22, the preprocessing unit 5, the image bus 22, the image memory 47, and so on, and the image data is sent and the preprocessing is repeated.

The image data that has undergone the preprocessing is sent to the characteristic part extraction section 6 and the characteristic part of the person is extracted. When the customer is classified by sex and customer data is created, the feature extraction unit 6 extracts the person by separating the person from the background,
Extraction of the face portion, clothing portion, etc. is performed. The image data of the extracted characteristic portion is stored in the image memory 4.

The image data of each characteristic portion is sent to the characteristic amount extraction unit 7. The feature amount extraction unit 7 extracts a feature amount for determining the gender based on the image data of each feature portion. As the feature amount for determining the gender, the body shape,
Hairstyle, chest size, whether or not makeup such as lipstick is applied, whether the button on the shirt is on the left front or right front, whether or not a skirt is worn, whether or not ornaments such as a necklace are attached, and the like.

Then, a score for sex determination is attached to each feature amount. In other words, the feature is classified into classes based on each feature amount, and the weight given in advance for each class is
The score of each feature amount is used. The score of each feature is determined by the determination unit 8
Sent to. The determination unit 8 refers to a knowledge base (dictionary) that stores a determination rule for a predetermined feature amount based on the score of each feature amount,
Make a comprehensive gender determination. This determination process may use a previously learned neural network instead of the knowledge base.

Information such as gender estimated in this way is
It is organized by the CPU unit 12 and recorded in the recording unit 10 such as a magnetic disk, or is transmitted from the communication control unit 11 to an external host via a communication line.

According to the above embodiment, it becomes possible to automatically collect the customer data which has been investigated by the human tactics until now. By linking the product name purchased by the vending machine and the customer data, it is possible to analyze the purchaser class of the product, so that the collected customer data can be effectively used for product planning work.

Although the customers are classified by sex in the above embodiment, the customers may be classified by age group. When classifying customers by age group, it is conceivable to use, for example, hairstyle, clothes, skin gloss, and the like as feature quantities. Alternatively, two video cameras may be installed to capture an image of a customer to generate a stereoscopic image, and various classifications may be performed based on the stereoscopic image.

Further, in the above embodiment, the video camera is installed in the vending machine, but it may be installed in the vicinity of the cash register of each booth at the department store. By doing so, it becomes possible to know the relationship between the product of each booth and the purchasing class of the product.

[0024]

According to the present invention, it is possible to automatically collect customer data, which has been investigated by human naval tactics, unattended.

[Brief description of drawings]

FIG. 1 is an electrical block diagram showing a configuration of a customer data automatic collection device.

[Explanation of symbols]

 1 Video Camera 2 A / D Converter 3 Frame Memory 4 Image Memory 5 Pre-Processing Unit 6 Feature Extraction Unit 7 Feature Extraction Unit 8 Judgment Unit 9 Knowledge Base Unit 10 Recording Unit 11 Communication Control Unit 12 CPU Unit

Continuation of the front page (51) Int.Cl. ⁶ Identification number Office reference number FI technical display location G06F 15/74 310 A 330 A

Claims (2)

[Claims] 1. An image pickup device for photographing a customer, a feature for extracting a feature of the customer by performing image processing on the picked-up image, and a classification for classifying the photographed customer into a predetermined category based on the extracted feature. An automatic customer data collection device comprising: means and a storage means for storing a classification result by the classification means. 2. An image pickup device for photographing a customer, an image memory for storing a customer image input from the image pickup device, and a pretreatment such as noise removal or feature enhancement for the customer image stored in the image memory. The processing means, the extraction means for extracting the characteristics of the customer from the pre-processed image, the classification means for classifying the customers into a predetermined category based on the extracted characteristics and the knowledge base, and the classification result. A customer data automatic collection device comprising a storage means for storing.