KR102473697B1 - Method for collecting data reflecting traveler's behavioral characteristics using facial recognition - Google Patents

Abstract

The present invention detects a traveler's face area from image information captured through a plurality of CCTVs installed in a plurality of areas of a tourist destination, selects coordinates consisting of characteristic points unique to each traveler from the detected face area, and converts the coordinate value into metadata. Extract and analyze the extracted metadata to derive information related to the age and gender of the traveler, assign a unique value to the metadata to generate traveler propensity information that calculates the ID for each traveler, and install it in multiple zones of the tourist destination From image information obtained from CCTV, tourist behavioral characteristic information including information on the tourist's visiting area related to metadata to which a unique value is assigned, visit time for each zone, and facial expression information for the corresponding zone is generated, and from the tourist behavioral characteristic information Provides a method for generating traveler behavioral characteristic data using facial recognition technology, which is characterized by calculating the stay time by zone, traffic line information in a tourist destination, and satisfaction by zone.

Description

Method for generating traveler's behavioral characteristics using facial recognition technology {Method for collecting data reflecting traveler's behavioral characteristics using facial recognition}

The present invention relates to a method and system for generating traveler behavioral characteristic data, and more particularly, proposes a new method for extracting traveler's propensity information and behavioral characteristic data using facial recognition technology and utilizing it for tourism services. do.

Travel brings a change to daily life and provides new experiences. It also provides opportunities for joy and growth for the young and old, and for romance and relaxation for the middle-aged. Before the popularization of travel, it was only an event held on special days such as marriage, graduation, and retirement. However, thanks to the development of transportation and transportation networks and the expansion of information and communication technology, travel has recently become a daily routine and daily life.

Travel agencies, tourist attraction management companies, accommodations, and restaurants develop travel products related to specific topics or specialized for each period according to changes in travelers' tastes and expansion of travel content, or provide new travel services that combine natural environments and experiences. . In order to promote these travel products or services, travel information is actively exposed using professional bloggers or portal sites rather than relying only on voluntary travel content producers. In addition, various information-oriented travel services are being developed, such as information on recommended seasonal destinations in consideration of the purpose of travel, in the provision of travel products centered on transportation and lodging.

Despite this trend, conventional travel-related services do not meet the trends of travelers. For example, it appears that travelers visiting Jeju Island visit and experience all areas of Jeju Island according to their personality and needs without being biased toward a specific tourist destination. It is only provided as a feedback, and it is extremely exceptionally limited to the extent of collecting surveys after travel.

All travel information that travelers want is different, and in order to improve the quality of travel, a high-level analysis of traveler tastes and objective travel facts is required. In particular, in order to provide smart tourism services for tourist destinations, it is important to understand the behavioral characteristics of tourists visiting tourist destinations. It is possible to receive daily visitor statistics through POS devices or big data services provided by local governments, but information that does not reveal the individual characteristics of travelers, such as time zone, age group, gender, and companion information for specific tourist destinations. It is only being provided. In this situation, in order to raise the level of tourism service and provide customized products for each traveler, it is necessary to establish a management strategy for each tourist destination or to deploy a separate staff and introduce a related system to analyze the characteristics of visitors. It is economically and technically difficult to obtain the information through investment.

Conventionally, there have been various attempts to increase the level of tourism service. For example, according to Patent Registration No. 10-1869023, information on the location of a guide drone flying in a tourist area and the location of a tourist is input to a repeater, Providing information input to the repeater to the central processing unit, analyzing the location and density of the tourists based on the information input to the central processing unit, and transmitting specific destination information input by the tourist through the transceiver terminal to the central processing unit. providing to a processing unit, and the central processing unit analyzes an optimal route to the specific destination based on the result of analyzing the location of tourists and the density of all tourists, and then transmits the information through at least one of the guide drone and the transceiver terminal. A method for guiding a tourist destination using a drone including guiding the optimal route has been proposed. Since this technology is based on the operation of drones and requires destination information entered through a tourist's terminal, there are many restrictions on its use by tourist attraction operators.

As another technology, there have been attempts to provide high-quality tourism services customized to each tourist and to develop various tourism products by utilizing big data. For example, various data such as portal and social search data, card usage data, bus usage data, and public wifi data for each tourist destination are collected to form big data and the results are utilized. For example, according to Registered Patent No. 10-2137653, the product sales rate according to the sales volume of a specific product group is calculated from the total sales volume of products sold in each region of the tourist destination, and the area with the high product sales rate is selected as the store location. By making recommendations, we propose a big data-based information provision system for the sales and development of tourist destination products that increases the sales of small and medium-sized merchants who sell tourist destination products and makes it easy to select a store location. In the case of this technology, analysis is done based on product payment information and traffic use information, and analysis of travelers' propensity and behavioral characteristics is not being performed.

In order to develop high-quality tourism services and smart tourism products, it is necessary to be able to easily check various statistical information on travelers who have visited tourist destinations. is needed In addition, information such as real-time visitor statistics, visitor characteristics analysis, visitor movement path or satisfaction analysis, and difficulties of the tourism weak that can be utilized by tourist destination companies are needed.

On the other hand, in an era when infectious diseases such as Corona 19 are prevalent, we monitor whether travelers are not concentrated in a specific area and maintain an appropriate number of people inside the tourist destination. It is necessary to operate tourist destinations that can minimize the spread of infectious diseases.

Conventionally, according to Patent Publication No. 10-2006-0021088, congestion information for each tourist destination is delivered to a mobile communication terminal through a mobile communication network to expand and provide better mobile communication services to subscribers, while knowing the degree of congestion for each tourist destination. This technology helps subscribers to modify their travel plans or change to less congested time slots. Since this technology provides tourist congestion information based on the use of mobile communication terminals, it is possible to maintain the number of people in each area within the tourist destination. It is not enough to use it to prepare real-time quarantine measures.

The present invention was conceived under the above-described technical background, and an object of the present invention is to provide a new method for analyzing characteristics such as gender, age, and presence or absence of companions of travelers visiting tourist destinations and deriving behavioral characteristics for them. will be.

Another object of the present invention is to provide a method for extracting objective information by which travelers who have visited a tourist destination can grasp their preferences for a specific area within the tourist destination.

Another object of the present invention is a method and system capable of extracting traveler propensity and behavioral information in a tourist destination so that consideration for the tourism weak among travelers and dispersal of personnel when an epidemic such as Corona 19 is widespread is to provide

Other objects and technical features of the present invention will be presented in more detail in the following detailed description.

In order to achieve the above object, the present invention detects the face region of a traveler from image information captured through a plurality of CCTVs installed in a plurality of areas of a tourist destination, and selects coordinates consisting of characteristic points unique to each traveler from the detected face region The corresponding coordinate value is extracted as metadata, the extracted metadata is analyzed to derive information related to the age and gender of the traveler, and a unique value is assigned to the metadata to generate traveler propensity information that calculates an ID for each traveler, Generating traveler behavioral characteristic information including information about the region visited by a traveler related to metadata to which a unique value is assigned, visit time for each region, and facial expression information for the region from image information obtained from CCTVs installed in multiple regions of the tourist destination; Provided is a method for generating traveler behavioral characteristic data using facial recognition technology, characterized in that the traveler's staying time by zone, movement information within a tourist destination, and satisfaction by zone are calculated from the traveler's behavioral characteristic information.

In the traveler propensity information, group type information including presence/absence of a traveler's companion and type of companion may be generated and additionally stored through metadata analysis obtained from image information captured in a plurality of zones.

In the present invention, in the case of the stay time, the time at which face information is confirmed for the visited areas is set as the visit time, and when moving from the first area to the second area, the second area visit time and the first area visit time and The stay time in the first zone can be determined from the difference excluding the average travel time in the corresponding tourist destination. In addition, in the case of the movement route information, the time at which face information is confirmed for the visited areas may be set as the visit time, and the movement route may be determined in the order of the visit time. In addition, in the case of the satisfaction, a facial expression conversion value obtained by numerically converting metadata related to the facial expression value is calculated, and the satisfaction value is calculated by the method of (number of visited areas) × (maximum value of numerical conversion values) - (total sum of facial expression conversion values) The value is calculated, and the closer the numerical value of satisfaction is to 0, the higher the satisfaction level is.

In the present invention, in addition to the face information of the traveler, the presence or absence of crutches, a wheelchair, or a stroller may be determined from the image information, and the corresponding information may be added to the metadata.

In addition, in the present invention, the number of travelers whose visiting time is confirmed for a specific area for a certain period of time is collected from the traveler behavioral characteristic information, and the number of travelers collected is compared to the standard number determined based on the area of the corresponding area to determine the density. can determine.

According to the present invention, the characteristics of travelers visiting a specific tourist destination can be grasped in real time and used to provide smart tourism services.

The present invention provides management strategy in individual tourist destinations by providing propensity information such as time zone, age group, gender, and companion characteristics of travelers visiting tourist destinations, and behavioral characteristic information such as visit area, stay time, and satisfaction within the tourist destination. Establishment or development of customized products for visitors can be actively carried out.

According to the method of the present invention, it is possible to accumulate traveler behavioral characteristic data of individual tourist destinations and use them for regional tourism industry analysis and policy establishment, and in the event of an epidemic such as Corona 19, the concentration of people in a specific area is monitored and the appropriate number of people is determined. It can also be used for real-time quarantine prevention activities such as dispersing.

1 is a flowchart of a method for generating traveler behavioral characteristic information according to the present invention
2 is a schematic diagram of tourist destination image information
3 is a schematic diagram of generating tourist propensity information after receiving video information
4 is a schematic diagram of generating traveler behavioral characteristic information
5 is a system configuration diagram for generating traveler behavioral characteristic information of the present invention
6 is a schematic diagram showing statistical information calculation results
7 is tourist destination analysis data transmitted to external institutions

The present invention proposes a method and system for generating traveler behavioral characteristic data using facial recognition technology.

The data generation method of the present invention obtains basic information from images captured through a plurality of CCTVs installed in tourist spots. Receives CCTV image information, detects the tourist's face area through facial recognition software, and extracts basic data such as gender, age group, and expression of each traveler from the detected face information to build a basic database. In this process, if the traveler includes a companion, the type of group may be classified based on the sex ratio and age information of the corresponding group.

In addition, in addition to the basic database, the present invention analyzes image information received from CCTVs in each area of the tourist destination to calculate traveler behavioral characteristic data including the traveler's stay time, traffic route, and satisfaction information for each area in the tourist destination. In addition, it is possible to determine the area of the area where CCTV is installed and the density of visitors within the area, and monitor whether the recommended number of people for quarantine is being followed. In each process, data reliability is improved by conducting a redundancy test with images taken by multiple CCTVs.

The tourist behavioral characteristic data calculated through the method of the present invention is statistical information such as real-time visitor aggregation, visitor characteristics analysis, companion analysis, satisfaction analysis, number of visitors by time slot, companion type, and difficulties of the tourism weak in tourist travel-related companies. can be utilized as In addition, agencies such as government offices that need to monitor or collect the number of local tourists are notified by hand of the number of visitors to each tourist destination, or the personnel in charge are directly surveyed. It can be used to establish or implement travel-related measures.

1 is a flow chart of a method for generating traveler behavioral characteristic information according to the present invention. The method for generating traveler behavioral characteristic data according to the present invention first includes the step of detecting a tourist's face region from image information captured through a plurality of CCTVs installed in a tourist destination. include

It is preferable to obtain the image information of the tourist from the CCTVs installed in the tourist destination from a plurality of CCTVs installed separately for each area (step S110). It can be, and the face region of each traveler is detected from the obtained image information. Thereafter, coordinates composed of characteristic points unique to each traveler are selected from the detected face area, and the corresponding coordinate values are extracted as metadata (step S120).

Figure 2 is a picture showing an example of image information of a tourist destination. Various data and statistical information can be obtained by using CCTV images installed in a tourist destination. It is desirable to delete the data without saving it.

It is necessary to determine the age and gender information of the traveler based on the result of analyzing the traveler's face area within the video information transmitted from the CCTV, and to calculate the group type and group type data by analyzing whether the visitor is an individual or multiple. . To this end, in the present invention, the metadata extracted from the face region of the image information is analyzed to derive information related to the age and gender of the traveler, and a unique value is assigned to the corresponding metadata to generate traveler propensity information that calculates an ID for each traveler. Do (step S130).

Metadata uses facial recognition technology to detect the facial area of all people from video images filmed through CCTVs installed in individual tourist destinations, select coordinates consisting of each person's unique feature points in the face area, and convert the corresponding coordinate value to meta data. It includes the process of extracting data. For example, by selecting the nose part of the face as a reference coordinate in the face area photographed by CCTV, modeling it in 3D around the reference coordinate, and detecting the facial feature point as a 3D coordinate value from the 3D face model. , individual metadata can be extracted.

Creation of traveler propensity information

Gender/age information of a person included in a photographed video object may be derived by comparing the extracted metadata with an existing accumulated age or gender metadata pattern. In the process of generating such traveler propensity information, multiple face regions can be detected through CCTVs for each region. can be upgraded or supplemented

In addition, in the process of generating traveler propensity information, a step of generating group type information including presence or absence of a traveler's companion and type of companion through metadata analysis obtained from image information captured in a plurality of zones may be included. It can be classified as an individual type or a group type through analysis according to the number of people detected in the recorded video, and when classified as a group type, the corresponding group type can be designated based on the gender/age information of each individual constituting the group. For example, group types can be classified as individuals, children, friends, lovers, parents, and organizations. In addition, in the case of friends, lovers, families, and other groups, each type may be classified according to the gender, age group, and age difference of the recognized person.

In this way, the traveler propensity information determined step by step from the face area of the image information is stored in the database after calculating the ID for each person by assigning a unique value to the metadata. FIG. 3 schematically shows the outline of generating traveler propensity information after receiving image information, which includes age, gender, and companion information of individual travelers. In the case of facial expression information, it may be reflected in the process of generating traveler behavioral characteristic information described later. have. In addition, Table 1 below shows the generated traveler propensity information as an example, and the metadata Traveler propensity information such as gender and age may be included.

ID metadata note 000001 data 000xxx1 individual 000002 data 000xxx2 married couple 000003 data 000xxx3 friend ...... 00000N data 000xxxN (group type)

tourism weak discrimination

The present invention is to determine the tourism weak before or during the process of generating tourist propensity information. As a step for this, it is possible to determine the existence of crutches, wheelchairs, and baby carriages in addition to the traveler's face information from the image information transmitted from the CCTV of the tourist destination, and add the corresponding information to the metadata (step S125). For tourism vulnerable information, for example, moving image information such as crutches, wheelchairs, strollers, and metadata information of pregnant women are learned, and when an image related to tourism vulnerable classification is found in the metadata for which group type classification has been completed, It is possible to differentiate between ordinary tourists and tourism weak by identifying travelers as tourism weak and adding the corresponding information to the corresponding metadata. Through this information extraction, it will be possible to identify the tourist weak who visits a specific tourist destination, while implementing measures such as changing the route and adding convenience facilities for the travelers.

Creation of information on traveler behavioral characteristics

Next, as a step of generating traveler behavioral characteristic information, using facial recognition technology, a traveler related to metadata to which a unique value is assigned is identified from image information obtained from CCTVs installed in multiple areas of a tourist destination, and information on the area visited by the traveler. , visit time for each zone, and tourist behavior characteristic information including facial expression information for the corresponding zone is generated (step S140). Even if it is a single tourist attraction, it is possible to check the visitor's movement route or stay time through multiple CCTVs installed in each area inside, which can be used as data to obtain information on stay time by section and verify the accuracy of group designation. .

4 is a schematic diagram for generating traveler behavioral characteristic information, schematically showing a process for measuring stay time, flow, and satisfaction through multiple CCTVs, such as CCTVs by area in a tourist destination and CCTVs at entrances and exits. Specifically, when a traveler with an assigned ID value is found on a CCTV in a specific area within a tourist destination, the information on the traveler's behavioral characteristics is confirmed by checking the area information, visit time, and expression status through facial recognition in the ID data. and include it as the corresponding ID and additional data. In the case of facial expression information, for example, it can be classified into four types of information such as Happy, Sad, Angry, and normal. Facial expression information can be digitized and used to calculate satisfaction. The same work is performed for each CCTV in each zone, and the top data is accumulated and stored, and when a tourist with the corresponding ID is found in the exit CCTV, the final exit zone value, visit time value, and facial expression value are given to update the data of the corresponding ID. quit

ID data for which renewal has been completed may be collected for each traveler and calculated as statistical information for all travelers visiting the tourist destination. Table 2 below shows, by way of example, information on regions visited by each traveler, visit time, and facial expression information for each region.

ID area visit time expression value
000001 A 12:10 normal B 12:35 happy C 13:35 happy D 14:20 sad
000002 A 13:20 happy C 14:30 normal D 14:50 normal 000003 B 11:00 happy C 12:30 normal
000004 A 14:25 sad B 14:40 happy D 15:30 normal

Next, the traveler's staying time for each zone, route information within the tourist destination, and satisfaction level for each zone are calculated from the traveler's behavioral characteristic information including the visit zone information, visit time for each zone, and facial expression information for the corresponding zone (step S150).

In the case of stay time, the time at which face information is confirmed for the visited areas is set as the visit time, and when moving from Area 1 to Area 2, the average visit time from Area 2 to Area 1 and the corresponding tourist destination The time of stay in Zone 1 can be determined from the difference except for the travel time. That is, a method of determining a stay time for a specific zone with a difference value of visit times for a plurality of zones can be expressed as follows.

(Visit time value for Zone 1)-(Visit time value for Zone 2)-(Average travel time) = Stay time

For example, if the final data of the tourist with the ID value of 000001 is as shown in Table 3 below, it can be seen that he moved from zone 1 through zone 3 to zone 2 in chronological order, and the average travel time is 10 min. The length of stay in the zone is derived as (11:20)-(10:30)-(00:10)=(00:40) according to the above formula. It can be determined that he stayed for 40 minutes and stayed for 50 hours in Zone 3 in the same way.

area visit time expression value travel time One 10:30 normal 8min 2 12:20 happy 12min 3 11:20 happy 10min

In the case of movement information, the time at which face information is confirmed for the visited areas is set as the visit time, and the movement may be determined in the order of the visit time. That is, in the final ID data value derived from the measurement of the residence time, the movement line can be measured by listing the zone values from the lowest data. For example, when the final data of a traveler with ID value of 000001 is expressed as a data set consisting of (area visited, visit time, facial expression information) (5,12:30,normal), (4,:11:20,normal ), (3,10:30,happy), and (1,10:10,happy), the traveler's movement can be easily determined as 1 -> 3 -> 4 -> 5.

In the case of the satisfaction, a facial expression conversion value obtained by numerically converting the metadata related to the expression value is calculated, and the satisfaction numerical value can be calculated by the following formula. In this case, the closer the satisfaction numerical value is to 0, the greater the satisfaction ( The closer to the number of visited areas) × (maximum value of numerical conversion value), the lower the level of satisfaction can be determined.

(number of areas visited)×(maximum value of numerical conversion value)-(total expression conversion value)

That is, satisfaction can be determined by listing the zone values and expression values from the lowest data in the final ID data value derived from the above stay time measurement. When the data is represented as a dataset consisting of (visit area, visit time, facial expression information) (5,12:30,normal), (4,:11:20,normal), (3,10:30,happy ), (1,10:10,happy), the traveler's satisfaction by zone can be identified as high in zone 1.3 and average in zones 4 and 5. In addition, the satisfaction of the traveler of the corresponding ID for the entire tourist destination can be derived by summing up the satisfaction of the individual zones and assigning a satisfaction score to the data of the facial expression value from the finally derived ID data. For example, calculate the total value converted numerically, such as happy = 4, normal = 3, sad = 2, angry = 1, and calculate ((number of visited areas) * 4) - (total expression conversion value) Through , it can be determined that the closer to 0, the higher the satisfaction, and the closer to (number of visited areas) * 4, the greater the dissatisfaction. In this way, it is possible to objectively analyze the behavioral characteristics of travelers in a specific tourist destination along with the above-mentioned stay time and movement data by summing up the satisfaction indices of individual travelers and statistically processing the satisfaction of all travelers who have visited the tourist destination.

After receiving the CCTV images, extracting metadata from the facial area to generate traveler propensity information, and generating information on each traveler's behavioral characteristics from the CCTV images by area, the total traveler data and the data by tourist destination calculate statistical data, It may be provided as visual information to a company or transmitted to a server of an external institution (step S160).

Density determination

In the present invention, the number of travelers whose visit time is confirmed for a specific area for a certain period of time is collected from the traveler behavioral characteristic information, and the density is determined by comparing the number of travelers collected against the standard number of people determined based on the area of the corresponding area. A determining step may be added (step S160).

In order to determine the tourist density by district, first of all, Based on the area of each zone, the value for the recommended number of tourists is determined and stored in the database. Regardless of the ID value for individual travelers, the total number of people in the video captured by CCTV in the area is identified, and the density of tourists in the area is identified in real time. If the result of (recommended number of tourists)-(current total number of people) is a negative number, the area concerned may judge that the negative number of tourists is exceeded and take precautionary measures. In this way, it is possible to monitor whether the proper density is maintained by identifying the real-time number of people compared to the area of each area of the tourist destination, and it is possible to operate a tourist destination that can minimize the spread of infectious diseases such as COVID-19 through the determination of density.

The method for generating traveler behavioral characteristic data of the present invention can perform the steps of extracting metadata and generating traveler-related information after receiving image information through wired/wireless communication at a place far from a tourist destination. can be configured. 5 shows an exemplary configuration of a system for generating traveler behavioral characteristic data of the present invention. The entire system 200 basically includes an image information collection unit 210, a metadata extraction unit 220, a tourist tendency information generation unit 230, and a traveler behavioral characteristic information generation unit 240, and additionally It may include a tourism weak information generation unit 235, a density determination unit 245, and a statistical information calculation unit 250. The image information collection unit receives image information from CCTVs for each area of the tourist destination, and the statistical information calculation unit may transmit and receive information with a server such as the tourism policy agency 300 through wired or wireless communication. Each component may be composed of independent circuit elements and parts capable of software processing, and functionally similar elements may be composed of one packaged part.

The traveler-related behavioral characteristic data generated according to the present invention is utilized by tourism-related companies to check various statistical information in real time, such as real-time visitor statistics, visitor characteristics analysis, companion analysis, satisfaction analysis, and difficulties of the tourism weak, and can be used to change services or develop specialized products. 6 schematically shows the result of calculating statistical information that can be utilized in a tourist destination according to the present invention.

In addition, if the statistical data of individual tourist destinations is comprehensively accumulated according to the present invention, it can be used as a basis for overall statistical analysis of the area or analysis of tourist behavior, and can be used to prepare quarantine measures for tourist destinations in real time. 7 schematically shows tourist destination analysis data transmitted to an external institution. Traveler-related behavioral characteristic data generated by each tourist destination is expected to greatly contribute to the development of tourism products in a specific area or to increase the level of local tourism.

Although the present invention has been illustratively described through preferred embodiments above, the present invention is not limited to such specific embodiments, and various forms are provided within the scope of the technical idea presented in the present invention, specifically, the claims. may be modified, changed, or improved.

100: CCTV by tourist area
200: Traveler behavioral characteristic data generation system
210: image information collection unit 220: metadata extraction unit
230: Traveler propensity information generation unit 235: Tourism weak information generation unit
240: Traveler behavioral characteristic information generation unit 245: Density determination unit
250: Statistical Information Calculation Unit 300: Tourism Policy Agency

Claims (5)

Detecting the tourist's face area from image information captured through multiple CCTVs installed in multiple areas of the tourist destination,
From the detected face area, coordinates consisting of unique feature points of each traveler are selected, and the corresponding coordinate values are extracted as metadata,
The extracted metadata is analyzed to derive information related to the age and gender of the traveler, and a unique value is assigned to the metadata to generate traveler propensity information that calculates an ID for each traveler.
Generating traveler behavioral characteristic information including information about the region visited by a traveler related to metadata to which a unique value is assigned, visit time for each region, and facial expression information for the region from image information obtained from CCTVs installed in multiple regions of the tourist destination;
From the traveler behavioral characteristic information, the traveler's stay time by zone, route information within the tourist destination, and satisfaction by zone are calculated,
The stay time is set as the visit time when the face information is confirmed for the visited areas, and when moving from the first area to the second area, the average movement from the second area visit time to the first area visit time and the corresponding tourist destination Determine the residence time in zone 1 from the difference excluding time,
In the case of the satisfaction, a facial expression conversion value obtained by numerically converting metadata related to the expression value is calculated, and a satisfaction numerical value is calculated by the following formula,
(number of areas visited)×(maximum value of numerical conversion value)-(total expression conversion value)
The closer the satisfaction numerical value is to 0, the greater the satisfaction, and the closer to (the number of visited areas) × (maximum value of the numerical conversion value), the lower the satisfaction. Characterized in that
A method for generating traveler behavioral characteristics data using facial recognition technology.
According to claim 1,
In the case of the traffic line information, the time at which face information is confirmed for the visited areas is set as the visit time, and the travel line is determined in the order of the visit time.
delete According to claim 1,
A method of generating traveler behavioral characteristic data using face recognition technology, characterized in that by determining the presence or absence of crutches, wheelchairs, and strollers in addition to face information of the traveler from the image information, and adding the corresponding information to the metadata.
According to claim 1,
The number of travelers whose visit time is confirmed for a specific area is collected from the traveler behavioral characteristic information for a certain period of time, and the density is determined by comparing the number of travelers collected to the standard number of people determined based on the area of the area A method for generating traveler behavioral characteristic data using facial recognition technology.

Images