KR20230008483A - Apparatus and method for measuring pedestrian volume - Google Patents

Abstract

Provided is a measuring device. The measuring device may comprise: a collection part that collects image data photographed in a set area; and a counting part that counts the pedestrians included in the image data. Therefore, the present invention is capable of automatically counting or measuring the number of pedestrians.

Description

Apparatus and method for measuring pedestrian volume}

The present invention relates to an apparatus and method for measuring a pedestrian traffic volume corresponding to a pedestrian traffic volume.

As society advances and develops, the leisure culture using walking rather than leisure activities using vehicles is increasing day by day. Although the floating population is increasing in downtown shopping streets, tourist attractions, parks, and hiking trails, it is not easy to determine how many people there are, when and where they are active.

In the past, surveyors measured the number of floating population by going to the site and counting by hand, but it is difficult to count while directly checking with human eyes. Because it is difficult to investigate for a long time, it was impossible to obtain reliable floating population data.

Therefore, recently, research on a pedestrian counting device capable of measuring the amount of pedestrian traffic in a specific area and a pedestrian counting management system using the same has been expanded.

Accordingly, in order to collect, analyze, and utilize the detected pedestrian traffic information, an additional device and method including a pedestrian traffic storage and transmission device is required.

Korean Patent Registration No. 1864221 discloses a technique for detecting pedestrian traffic using a PIR sensor and collecting and monitoring data.

Korean Registered Patent Publication No. 1864221

An object of the present invention is to provide a measuring device and method for measuring the amount of walking and generating a map based on the amount of walking.

The measurement device of the present invention includes a collection unit for collecting image data taken in a set area; A counting unit counting the number of pedestrians included in the image data may be included.

The measurement method of the present invention includes a collection step of collecting image data obtained by photographing a set area; An extraction step of extracting a pedestrian through analysis of the image data; a counting step of counting the pedestrian included in the image data; and a transmission step of providing a counting result of the pedestrian to a server generating a walking amount map.

According to the present invention, the number of pedestrians using the corresponding walking path may be automatically counted or measured using video data of the walking path captured by a camera such as CCTV.

According to the present invention, a human object corresponding to a pedestrian included in image data of a walking path may be extracted, and a walking amount may be calculated. The calculated walking amount may be provided to a server capable of creating a walking amount map together with place information indicating a target point of the walking amount.

The walking amount map generated by the measuring device of the present invention may display walking amount information in real time. For example, by using a walking amount map, various walking characteristics, such as the priority of creating a walking environment and where the walking amount is high or low, can be analyzed.

According to the present invention, in the current situation where it is difficult to accurately estimate the amount of walking and it is difficult to know how much the amount of walking is at any location, a walking amount map can be provided by which anyone can obtain information on the amount of walking.

1 is a schematic diagram showing a measuring device of the present invention.
Figure 2 is a schematic diagram showing the operation of the extraction unit and the counting unit.
Figure 3 is a schematic diagram showing the operation of the measuring device of the present invention.
4 is a schematic diagram showing a walking volume map generated by a server;
5 is a flowchart showing the measurement method of the present invention.
6 is a diagram illustrating a computing device according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

In this specification, redundant descriptions of the same components are omitted.

In addition, in this specification, when a component is referred to as being 'connected' or 'connected' to another component, it may be directly connected or connected to the other component, but another component in the middle It should be understood that may exist. On the other hand, in this specification, when a component is referred to as 'directly connected' or 'directly connected' to another component, it should be understood that no other component exists in the middle.

In addition, terms used in this specification are only used to describe specific embodiments and are not intended to limit the present invention.

Also, in this specification, a singular expression may include a plurality of expressions unless the context clearly indicates otherwise.

In addition, in this specification, terms such as 'include' or 'having' are only intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more It should be understood that the presence or addition of other features, numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Also in this specification, the term 'and/or' includes a combination of a plurality of listed items or any item among a plurality of listed items. In this specification, 'A or B' may include 'A', 'B', or 'both A and B'.

Also, in this specification, detailed descriptions of well-known functions and configurations that may obscure the subject matter of the present invention will be omitted.

Currently, the government has a high interest in creating a pedestrian environment. In order to create and plan a walking environment, it is necessary to precede the task of calculating the amount of walking. Currently, only an approximate amount of walking can be calculated through location data of individual mobile phones possessed by the user, and it is not possible to know exactly how many walking paths 50 there are. In order to solve this problem, the measurement device 100 of the present invention may collect image data captured by the CCTV 10 on the walking path 50 or image data of various terminals 30 photographing the walking path 50. The measurement device 100 may detect an object of a person walking on the walking path 50 from image data and make it a walking amount. The measuring device 100 may transmit pedometer information to the server 90 together with place information including a location where the image data was captured. The server 90 may generate a walking amount map corresponding to a map representing the walking amount by using the place information and the walking amount information.

If such a walking amount map is provided, it is expected that various walking characteristics such as the priority of creating a walking environment and the high or low amount of walking at a certain location can be analyzed.

1 is a schematic diagram showing a measuring device 100 of the present invention.

The measurement device 100 shown in FIG. 1 may include a collection unit 110 , an extraction unit 130 , a counting unit 150 , and a transmission unit 170 .

The collecting unit 110 may collect image data obtained by photographing a set area.

For example, the collection unit 110 may communicate with a closed circuit TV (CCTV) 10 that captures a set area. Collecting unit 110 may obtain image data from the CCTV (10).

The collection unit 110 may acquire a plurality of image data from a plurality of regions in real time. Numerous image data obtained in real time are processed by the extraction unit 130 and the counting unit 150, and in this process, an enormous processing load may be applied to the extraction unit 130 or the counting unit 150.

In order to reduce the processing load of the extraction unit 130 and the counting unit 150, a pre-processing unit 19 may be provided between the CCTV 10 and the collecting unit 110. For example, the pre-processing unit 19 may be installed in the CCTV (10).

The pre-processing unit 19 may lower the resolution of the image data to a minimum resolution capable of recognizing a person (pedestrian) corresponding to an object included in the image data. Through downsizing, which lowers the resolution, the capacity of image data can be reduced to 30 to 40% compared to the original. The collection unit 110 may receive image data downsized by the pre-processing unit 19 .

The extraction unit 130 may extract a pedestrian through analysis of image data.

The extraction unit 130 may extract a pedestrian from image data using You Only Look Once (YOLO), which is a deep learning object detection technique.

The counting unit 150 may count pedestrians included in the image data.

The counting unit 150 may count the pedestrians extracted by the extracting unit 130 . For example, when the pedestrian extracted by the extraction unit 130 passes through a specific point in the image data, the counting unit 150 may increase the number of counts corresponding to the result of counting the pedestrian. In other words, the counting unit 150 may increase the count by one whenever a pedestrian passes a specific point in the image data.

2 is a schematic diagram showing operations of the extraction unit 130 and the counting unit 150.

The extractor 130 may extract a pedestrian included in the image data using an object detection technique such as YOLO and generate a virtual rectangle t inscribed with the pedestrian.

The counting unit 150 may form a straight line virtual count line d in the image data.

When the rectangle t moving along the pedestrian passes the count line d, the counting unit 150 may increase the number of counts corresponding to the result of counting the pedestrian.

At this time, when a point of the rectangle t reaches the count line d, a problem of increasing the count number may arise.

The amount of pedestrians may mean the number of pedestrians passing through a set point. In this case, when the center of gravity of the pedestrian reaches the set point, it would be reasonable to regard it as having passed the set point.

The extraction unit 130 may generate a rectangle t to have a side k2 perpendicular to the count line d.

The counting unit 150 may increase the number of counts when the half point km of one side k2 passes through the count line d. The counting unit 150 may increase the count number when the half point km reaches the count line d after one end of one side k2 reaches the count line d. The half point km of one side k2 may correspond to a midpoint point of one side k2.

Meanwhile, in order to accurately measure the amount of walking, it is preferable to provide a method for preventing the count number from being increased when a specific pedestrian walks back and forth on the count line d.

For example, the counting unit 150, once the count number increases due to a specific rectangle reaching the half point km on the count line d, further increases the count number due to the specific rectangle until the entire side k2 deviates from the count line d. can suppress In other words, even if one end of a particular rectangle enters the count line d and a pedestrian walks in a state where a part of one side k2 of the rectangle crosses the count line d, the count number is not further increased.

The display screen of the image data shown in FIG. 2 shows the walking path 50 of the set area.

In (a) of FIG. 2 , a pedestrian appears on the right side of the walking path 50 . The extraction unit 130 may extract a pedestrian appearing on the display screen using various object detection techniques and may cover the pedestrian with a virtual rectangle t inscribed with the pedestrian. The corresponding rectangle t can move along with the pedestrian.

A count line d set by the counting unit 150 may be formed on the display screen. The count line d may be set by an administrator in advance.

One side k2 of the rectangle t may be perpendicular to the count line d. The other side k1 of the rectangle t perpendicular to one side k2 may be parallel to the count line d.

In (a) of FIG. 2 , since there is no rectangle t passing through the count line d, the number of detected pedestrians 'Detected Pedestrians' is 0.

2(b) shows a state in which two pedestrians appearing from the right side of the walking path 50 have passed the count line d. A part of the rectangle t inscribed by one of the pedestrians still crosses the count line d, but the half point km of one side k2 of the rectangle t passes the count line d, so the count number may be increased. Since a total of 2 pedestrians passed the count line, the number of detected pedestrians 'Detected Pedestrians' is 2.

Meanwhile, a method for improving the accuracy of measuring the amount of walking in a set area may be prepared.

For example, the measuring device 100 of the present invention may measure the amount of walking by using a plurality of image data obtained by photographing different points in a set area.

3 is a schematic diagram showing the operation of the measuring device 100 of the present invention.

A plurality of CCTV (10) cameras to photograph different points in the set area may be provided. In this case, the collection unit 110 may collect at least one of a plurality of image data obtained by photographing different points in the set area and location information of the terminal 30 carried by the pedestrian.

At this time, the counting unit 150 may exclude the location information of the terminal 30 and calculate the amount of pedestrians in the set area using the number of pedestrians counted at different points.

Alternatively, the counting unit 150 may calculate the amount of pedestrians in the set area by using the number of pedestrians counted at a specific point and the location information of the terminal 30 together.

Alternatively, the counting unit 150 may calculate the amount of pedestrians in the set area by using the number of pedestrians counted at different points and location information of the terminal 30 together.

When the location information of the terminal 30 is obtained in real time, the number of terminals 30 that have passed through the set area or the count line d can be recognized in real time. However, the counting method using location information may have a limitation in that it counts the number of terminals 30 possessed by the pedestrian rather than actually counting the number of pedestrians. Accordingly, the location information may be used for the purpose of identifying an error existing in the number of pedestrians counted through image data.

Since there may be pedestrians who do not carry the terminal 30, the number of location information corresponding to the terminal 30 on a one-to-one basis may be less than or equal to the number of pedestrians. Therefore, the counting unit 150 may determine that an abnormal situation has occurred if the number of pedestrians confirmed through the image data is smaller than the number of terminals 30 counted through the location information.

The abnormal situation at this time may indicate at least one of an error of the CCTV 10 capturing the set area, an error of the collection unit 110, and an error of the counting unit 150.

The collecting unit 110 may collect first image data i1 obtained by capturing a first point p1 of the set region and second image data i2 obtained by capturing a second point p2 of the set region. The first image data i1 may be acquired from the first CCTV 11 capturing the first point p1. The second image data i2 may be acquired from the second CCTV 12 capturing the second point p2.

The counting unit 150 may calculate the amount of pedestrians in the set area using the first result of counting the pedestrians included in the first image data i1 and the second result of counting the pedestrians included in the second image data i2.

The counting unit 150 may determine the average value of the first result and the second result as the amount of walking in the set area.

The first point p1 may include the starting point of the setting area. The second point p2 may include an end point of the setting region.

For example, if the length w of the set region is 100 m, the first image data i1 may be taken at a point 0 m based on the set region. The second image data i2 may be taken at a point 100m away from the set area.

The counting unit 150 may analyze the first image data i1 and the second image data i2 for a set time period and count pedestrians in each image data.

As a result of counting, 100 pedestrians passed through the first count line d1 formed at the first point p1 and 100 pedestrians passed through the second count line d2 formed at the second point p2. In this case, the counting unit 150 may measure or calculate an average of 100 pedestrians in the set area during the set time period.

Meanwhile, in a state where 95 pedestrians pass through the first count line d1, 90 pedestrians pass through the second count line d2, and the remaining 5 pedestrians may return to the first point and pass through the first count line d1 again. In this case, pedestrians passing through the first count line d1 may be counted as 100, and pedestrians passing through the second count line d2 may be counted as 90. According to the present invention, since the average value of the first result and the second result is used, 95 people, which is the average value of 90 people and 100 people, can be measured and calculated as the number of pedestrians or the amount of walking.

The transmission unit 170 may provide the pedestrian counting result and location information together to the server 90 that generates the walking amount map. The location information is information acquired by the collection unit 110 and may include location information where the image data was captured.

The server 90 may specify a point or a section of the map using the place information provided from the transmission unit 170, and may display the amount of pedestrians at the specified point or section using the counting result. In this way, a walking amount map corresponding to the map on which the walking amount is displayed can be generated by the server 90 and provided to a user who needs a walking amount route.

4 is a schematic diagram showing a walking volume map generated by the server 90 .

4 may show walking paths 50 in some areas included in the map.

The server 90 may classify and display the walking amount through color classification on the walking path 50 displayed on the map. A map in which the walking amount is separately displayed may correspond to the walking amount map m.

For example, the server 90 may display an area a with a high pedestrian traffic among the walking paths 50 in a first color, an area b with a moderate pedestrian traffic in a second color, and an area c with a low pedestrian traffic in a third color. there is.

5 is a flowchart showing the measurement method of the present invention.

The measuring method shown in FIG. 5 may be performed by the measuring device 100 shown in FIG. 1 .

The measurement method may include a collection step (S 510), an extraction step (S 520), a counting step (S 530), and a transmission step (S 540).

In the collecting step (S510), image data of the set area and location information including the location where the image data was captured may be collected together. The collecting step (S510) may be performed by the collecting unit 110.

In the extraction step (S520), pedestrians may be extracted through analysis of image data. The extraction step (S520) may be performed by the extraction unit 130. The extractor 130 may extract pedestrians included in the image data using the extraction model.

The extraction model may be machine-learned through a separate learning unit. The learning unit may perform machine learning on people (objects) included in the image data through YOLO v5.

Using the characteristic that pedestrians generally travel in only one direction when passing through the walking path 50, the learning unit uses two CCTVs (10) (video data at the first point and video data at the second point) and a pedestrian's terminal (30) (acquisition of location information in units of 1 second) can be used to increase the accuracy of object detection.

In addition, when non-human objects (animals, etc.) are included in the image data, in the last detecting part of the Yolo algorithm, all non-human objects receive only false values, and extraction Unit 130 may not perform object detection for this.

The extraction model generated by the learning unit can be trained through hundreds of thousands of human images, and a human object can be accurately identified with a learning rate of 95% or more, for example, 98.24%.

In order to approach the probability of 100% in addition to the learning rate learned in YOLO, the counting unit 150 mixes the amount of objects detected by the first CCTV 11 and the amount of objects of the second CCTV 12 on a continuous line It is possible to calculate the exact number of object detections through the object detection comparison value by performing the analysis.

In addition, when the location information of the terminal 30 is used, the coordinates and travel time of the pedestrian carrying the terminal 30 are expressed, and the amount of objects detected by the CCTV 10 is compared with the value transmitted by the terminal 30. After analysis, the amount of objects that come within the 5% range can be calculated as the final object detection amount.

The algorithm of this process is repeatedly calculated, and finally, an extraction model is constructed by fitting a model for object detection through a genetic algorithm in which only good algorithms remain.

In the counting step (S530), pedestrians included in the image data may be counted. The counting step (S530) may be performed by the counting unit 150.

The amount of walking for a corresponding section (set area) can be calculated using the amount of objects identified through the extraction model.

In the transmission step (S540), the pedestrian counting result and location information may be provided together to the server 90 that generates the walking amount map. The transmission step (S540) may be performed by the transmission unit 170. The server 90 may match location information including location information of the CCTV 10 and a shooting point to the amount of walking and display the location information on a map.

Basically, as shown in Table 1, when estimating the pedestrian service level from the road capacity manual, the server 90 can calculate the pedestrian convenience service level if only the pedestrian volume and the specifications of the corresponding walking path 50 are met.

A is the best condition, and the lower the English alphabet, the worse the condition.

When evaluating the service level, the specifications for the facilities of the pedestrian path 50 may be fixed values from the time of design and construction.

The amount of walking may be a non-fixed value or a floating value that changes over time.

Through the real-time foot traffic calculated in the present invention, the service level can be visualized for each walking path 50 by time zone, by day, by peak, and by non-peak. The 'peak star' may represent a peak value of the amount of walking. 'Non-peak star' may indicate a normal or average amount of walking.

Service level can be related to walking comfort. For example, even if the amount of walking is large, if the walking path 50 is wide, the pedestrian may not feel any discomfort while walking. In this case, the service level can be highly evaluated. Even if the walking amount is small, if the walking path 50 is narrow, the pedestrian may feel uncomfortable while walking. In this case, the service level may be evaluated as low.

In addition, as a method of recognizing the risk of spreading various viruses and diseases, the server 90 may select an area with a low service level as an infection risk area and add it to the pedestrian traffic map.

service level Pedestrian traffic rate
(person/minute/m) occupied space
(㎡/person) density
(Persons/㎡) speed
(m/min) A ≤20 ≥3.3 ≤0.3 ≥75 B ≤32 ≥2.0 ≤0.5 ≥72 C ≤46 ≥1.4 ≤0.7 ≥69 D ≤70 ≥0.9 ≤1.1 ≥62 E ≤106 ≥0.38 ≤2.6 ≥40 F - < 0.38 > 2.6 < 40

6 is a diagram illustrating a computing device according to an embodiment of the present invention. The computing device TN100 of FIG. 6 may be a device described in this specification (eg, the measuring device 100, etc.).

In the embodiment of FIG. 6 , the computing device TN100 may include at least one processor TN110, a transceiver TN120, and a memory TN130. In addition, the computing device TN100 may further include a storage device TN140, an input interface device TN150, and an output interface device TN160. Elements included in the computing device TN100 may communicate with each other by being connected by a bus TN170.

The processor TN110 may execute program commands stored in at least one of the memory TN130 and the storage device TN140. The processor TN110 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. Processor TN110 may be configured to implement procedures, functions, methods, and the like described in relation to embodiments of the present invention. The processor TN110 may control each component of the computing device TN100.

Each of the memory TN130 and the storage device TN140 may store various information related to the operation of the processor TN110. Each of the memory TN130 and the storage device TN140 may include at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory TN130 may include at least one of read only memory (ROM) and random access memory (RAM).

The transmitting/receiving device TN120 may transmit or receive a wired signal or a wireless signal. The transmitting/receiving device TN120 may perform communication by being connected to a network.

Meanwhile, the embodiments of the present invention are not implemented only through the devices and/or methods described so far, and may be implemented through a program that realizes functions corresponding to the configuration of the embodiments of the present invention or a recording medium in which the program is recorded. And, such implementation can be easily implemented by those skilled in the art from the description of the above-described embodiment.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concept of the present invention defined in the following claims are also provided. fall within the scope of the invention.

10... CCTV 11... 1st CCTV
12... 2nd CCTV 19... pre-processing unit
30 ... Terminal 50 ... Walkway
90 ... server 100 ... measuring device
110 ... collection unit 130 ... extraction unit
150 ... counting unit 170 ... transmission unit

Claims (10)

a collection unit that collects image data taken in a set area;
a counting unit counting the number of pedestrians included in the image data;
Measuring device comprising a.
According to claim 1,
The collection unit communicates with a CCTV (Closed Circuit TV) for photographing the set area,
The collection unit measuring device for acquiring the image data from the CCTV.
According to claim 1,
An extraction unit for extracting the pedestrian through analysis of the image data is provided;
Wherein the counting unit increases a count number corresponding to a pedestrian counting result when the pedestrian extracted by the extracting unit passes a specific point in the image data.
According to claim 1,
An extractor for extracting the pedestrian from the image data using YOLO (You Only Look Once), a deep learning object detection technique, is provided;
The counting unit counts the pedestrians extracted by the extraction unit.
According to claim 1,
An extraction unit for generating a virtual rectangle in which the pedestrian included in the image data is inscribed is provided;
The counting unit forms a linear virtual count line in the image data;
Wherein the counting unit increases a count number corresponding to a result of counting the pedestrian when the rectangle moving along the pedestrian passes through the count line.
According to claim 5,
The extraction unit generates the rectangle to have one side perpendicular to the count line,
The counting unit increases the count number when a half point of the one side passes through the count line.
According to claim 6,
The counting unit increases the count number when the half point reaches the count line after one end of the one side reaches the count line;
When the counting unit increases once due to the specific rectangle at the half point of the count line, suppressing further increase in the count number due to the specific rectangle until all of the one side deviate from the count line Measurement Device.
According to claim 1,
The collecting unit collects at least one of a plurality of image data captured at different points in the set area and location information of a terminal carried by the pedestrian;
The counting unit excludes the location information and calculates the amount of pedestrians in the set area using the number of pedestrians counted at different points,
The counting unit calculates the amount of pedestrians in the set area by using the number of pedestrians counted at a specific point together with the location information;
Wherein the counting unit calculates the number of pedestrians counted at different points and the location information together to calculate the amount of pedestrians in the set area.
According to claim 1,
The collecting unit collects first image data obtained by photographing a first point in the set region and second image data obtained by photographed a second point in the set region;
The counting unit calculates the amount of pedestrians in the set area using a first result of counting pedestrians included in the first image data and a second result of counting pedestrians included in the second image data;
The counting unit determines the average value of the first result and the second result as the amount of walking in the set area.
In the measuring method performed by the measuring device,
A collection step of collecting image data of a set area and location information including a location where the image data was captured together;
An extraction step of extracting a pedestrian through analysis of the image data;
a counting step of counting the pedestrian included in the image data;
a transmission step of providing the pedestrian counting result and the place information together to a server generating a walking amount map;
A measurement method comprising a.

Images