WO2023035553A1

WO2023035553A1 - Monitoring method and apparatus for service window and self-service device

Info

Publication number: WO2023035553A1
Application number: PCT/CN2022/077303
Authority: WO
Inventors: 邓小飞; 江晨; 程求红; 孙非凡; 周院平; 唐红; 孙信中; 矫人全
Original assignee: 南京奥拓电子科技有限公司
Priority date: 2021-09-09
Filing date: 2022-02-22
Publication date: 2023-03-16
Also published as: CN113766191A

Abstract

The present invention belongs to the field of video monitoring. Disclosed are a monitoring method and apparatus for a service window and a self-service device. The method comprises: detecting whether there is a person in a monitoring area; when it is detected that there is a person in the monitoring area, acquiring information of the person by means of facial recognition and/or portrait feature recognition, and establishing an identity identifier according to the information; and performing monitoring video recording on the monitoring area, performing grouped storage on a recorded video according to the identity identifier, and storing, in the same group, videos recorded under the same identity identifier. Therefore, the monitoring situations of a service window and a self-service device are efficiently recorded; the storage quantity of invalid videos is reduced, thereby improving the storage efficiency; the grouped storage of videos facilitates a later query of a monitoring video that is required by a client; and a recorded video can be quickly positioned according to an identity identifier.

Description

Method and device for monitoring business window and self-service equipment

technical field

The invention relates to the field of video monitoring, and more specifically, to a monitoring method and device for business windows and self-service equipment.

Background technique

At present, self-service equipment and business window video surveillance systems generally monitor video recording around the clock. When there is no one in the monitoring area, the system is also recording surveillance video, and the recorded video has almost no practical value. If there is no one in the self-service equipment monitoring video system for a long time or no one is handling business at the outlet business window for a long time, and it is idle, then the data in the idle state is almost useless, thus monitoring invalid video data, resulting in too much storage Invalid data increases the storage space of the video storage device, and also makes the amount of data stored in the video storage device too large.

Since invalid video data takes up a large part of the storage space, if a dedicated mass storage device is used to store the surveillance video recorded by the self-service device, the overhead of the system will be increased. Of course, if the surveillance video recorded by the self-service equipment is stored in the way of circular storage, after the storage device is full, the newly generated data will overwrite the previous data, shortening the video storage time. When querying the surveillance video required by the customer, if the required surveillance video has been covered by the newly generated video data, the required surveillance video will not be queried, which may cause losses to the customer.

There are also solutions to the problem of invalid video data occupying storage space in the prior art, for example: Chinese patent: a video surveillance processing method and device for self-service equipment, publication number: CN103049955B publication date: 2015-09-23, discloses a A self-service equipment video monitoring processing method, the method comprising: detecting whether there is a person in the self-service area; in the case of detecting that there is a person in the self-service area, triggering the self-service equipment monitoring video system to record the monitoring video, and storing the self-service equipment monitoring video The monitoring video recorded by the system; when no one is detected in the self-service area, the self-service equipment monitoring video system stops monitoring video recording. Therefore, in this application, when there is no one in the self-service area, the self-service equipment video surveillance system will not perform surveillance video recording. Effectively, reduces storage space. Due to the reduction of invalid video data, the workload of querying video records is also reduced.

However, the existing technology can only detect whether there is someone in the self-service area, and cannot detect the identity of the person. If an employee of the branch enters the self-service area, it will also trigger surveillance video recording, resulting in invalid video recording, which takes up a lot of storage Space will still increase the cost of storage devices.

Contents of the invention

1. Technical problems to be solved

Aiming at the problem of invalid video data occupying storage space in the prior art, the present invention provides a monitoring method for business windows and self-service equipment, which uses face recognition and portrait feature recognition to monitor the personnel who enter the self-service equipment and business windows Identity identification is made, and video storage is grouped and stored based on each identity identification, so as to efficiently record the monitoring situation of business windows and self-service equipment, reduce the amount of invalid video storage, and improve storage efficiency.

2. Technical solution

The purpose of the present invention is achieved through the following technical solutions.

A method for monitoring a business window and self-service equipment, comprising the following steps:

Step 1: Record the monitoring video images of the branch business window and the self-service equipment area through the network camera, as the input source for the subsequent steps;

Step 2: Obtain the monitoring video image collected in step 1, and select the area range that needs to be monitored in the monitoring video image;

Step 3: Detect whether there is someone in the framed area, if there is, identify the person's face information, establish a corresponding identity for the person based on the face information, and save the face information and the corresponding identity;

Step 4: If someone is detected in the framed area in step 3, record the surveillance video of the framed area, and save the recorded video as a group, which is stored based on the identity of the person, and only for valid videos Store them to avoid storing invalid videos, and store them in groups based on identification, which is conducive to backtracking after an event occurs.

Furthermore, if no one is detected in the framed area in step 3, then in step 4, the framed area is monitored for video recording, but the recorded video is not stored, and invalid video data can be avoided by not storing the recorded video Take up storage space.

Furthermore, 3 parameters are preset: detection period, alarm period and consecutive hit times. In step 3, the detection period is used as the time interval to periodically detect whether there is anyone in the frame selection area. If there is someone, it is considered a hit. In an alarm cycle, if the total number of hits reaches the number of consecutive hits, it is considered that there are people in the frame selection area. These three parameters can control the recall rate and precision rate of the detection method. In the actual detection, the parameters are adjusted according to the actual situation, so as to obtain the best detection results.

Furthermore, in step 3, the self-service equipment monitoring area is framed for the monitoring video image of the self-service equipment, and the outlet staff working area and customer monitoring area are respectively framed for the monitoring video image of the business window.

Furthermore, step 3 also includes:

Compare the person's face information with the saved face information, if the person's face information is the same as a saved face information, set the person's identity as the same saved face information The saved face information corresponds to the saved identity, thus avoiding the establishment of two or more identity marks for the same person, ensuring that the identity is unique, and the video of the same person can also be stored when storing videos in the same group.

Furthermore, if no face information is recognized in step 3, identify the person's portrait features, establish a corresponding identity for the person based on the portrait features, and save the portrait features and the corresponding identity.

Furthermore, identifying the features of the portrait includes: the specific method of identifying the features of the portrait is: identifying one or more of hair color, hair length, clothing color, clothing pattern, height, and body shape.

Furthermore, step 3 also includes:

Compare the person's portrait features with the saved portrait features, if the person's portrait features are the same as a saved portrait feature, then set the person's identity as the saved face information with the same portrait features The corresponding saved identity.

Furthermore, in step 3, if it is detected that there is someone in the framed area, the face information of the identified person is compared with the face information of the branch staff to determine whether the person is a branch employee, if the person is not a branch employee , the person is considered a customer;

If it is detected in step 3 that there are people in the monitoring area of the self-service equipment, record the monitoring video in the monitoring area of the self-service equipment, and save the recorded video as a group, which is stored based on the identity of the person;

If it is detected in step 3 that there is no one in the monitoring area of the self-service equipment, record the monitoring video of the monitoring area of the self-service equipment.

If it is detected that there is someone in the customer monitoring area in step 3, and the person is identified as a customer, then the monitoring video of the customer monitoring area is recorded, and the recorded video is saved as a group, which is stored based on the identity of the person ;

If it is detected in step 3 that there are people in the work area of the branch personnel, and the person is identified as a branch employee, then the monitoring video recording is performed on the work area of the branch personnel.

Another aspect of the present invention provides a monitoring device for business windows and self-service equipment, which is used to implement the above-mentioned monitoring method for business windows and self-service equipment, including:

The input module uses the monitoring video images entered in the business window of the network point and the self-service equipment area as the input source for the subsequent steps;

The frame selection module is used to frame the area range that needs to be monitored in the surveillance video image;

The detection module is used to detect whether there is a person in the frame selection area, and if there is a person, identify the face information of the person, establish a corresponding identity mark for the person according to the face information, and save the face information and the corresponding identity mark;

The recording and saving module is used to record the monitoring video of the framed area if a person is detected in the framed area in step 3, and save the recorded video as a group, which is stored based on the identity of the person.

3. Beneficial effect

Compared with the prior art, the present invention has the advantages of: efficiently recording the monitoring situation of business windows and self-service equipment, reducing the storage capacity of invalid videos, and improving storage efficiency; it is convenient to check the video monitoring of self-service equipment and business windows in real time; The personnel of the equipment and the business window make identities, and store them in groups based on each identity, which is convenient for later review through surveillance video, improves service quality, reduces customer dissatisfaction rate, and increases customer stickiness.

Description of drawings

Fig. 1 is the overall flow chart of the present invention.

Detailed ways

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

Example 1

The embodiment of the present invention provides a monitoring method for a business window and self-service equipment, as shown in Figure 1, the process of the method is as follows:

Step 1: Record the monitoring video images of the business window of the outlet and the self-service equipment area through the network camera, as the input source for the subsequent steps. The camera used in this embodiment is a camera that supports the RTSP standard protocol and the H.264 encoding format, and its maximum resolution is 1080P.

Step 2: Obtain the monitoring video image collected in step 1, and frame the range of the area to be monitored in the monitoring video image, so as to detect the framed area later, and the framed area can be selected according to actual needs.

Step 3: Detect whether there is anyone in the framed area. If there is someone, identify the person's face information, establish a corresponding identity for the person based on the face information, and save the face information and the corresponding identity; In the case of face information, identify the portrait features of the person, establish a corresponding identity mark for the person according to the portrait features, and save the portrait features and the corresponding identity mark. The face recognition technology in this step is an existing technology and can be completed by a face recognition technology provided by a third party. Face recognition is a biometric technology for identification based on human facial feature information. A technology that collects images or video streams containing human faces with cameras or cameras, automatically detects and tracks human faces in the images, and then performs identity verification or identity search for detected faces. Face recognition can recognize information such as a person's name, age, and face picture, and use this information as an identity. By adjusting the threshold of the face recognition algorithm, a better pass rate and a lower false recognition rate can be obtained. Portrait features include: hair color, hair length, clothing color, clothing pattern, height and body shape, when facial information cannot be recognized. Personnel can be identified through portrait features.

In order to ensure the uniqueness of the identification, compare the person's face information/portrait features with the saved face information/portrait features, if the person's face information/portrait features are compared with a saved face information / portrait features are the same, then the identity of the person is set to the saved identity corresponding to the saved face information/portrait features that are the same as its face information, thereby avoiding the establishment of two and two images for the same person. More than one identity mark, when storing videos, the videos of the same person can also be stored in the same group.

In order to detect whether there are people in the frame selection area, three parameters are preset: detection cycle, alarm cycle, and consecutive hit times. With the detection cycle as the time interval, it is periodically detected whether there is anyone in the frame selection area. If there is someone, it is considered a hit , within an alarm period, if the total number of hits reaches the number of consecutive hits, it is considered that there are people in the frame selection area. These three parameters can control the recall rate and precision rate of the detection method. During the actual detection, the parameters are adjusted according to the actual situation, so as to obtain the best detection results.

Step 4:

If it is detected in step 3 that there are people in the framed area, the framed area is monitored and video recorded, and the recorded video is saved as a group. The group is stored based on the identity of the person, and only valid videos are stored. Avoid storing invalid videos, and store them in groups based on identity identification, which is conducive to backtracking after an event occurs. The identification includes: name, age and face picture. In this embodiment, the name is used to name the video storage during storage, so that the video content can be quickly located according to the identification when querying the video monitoring required by the customer later. .

If no one is detected in the framed area in step 3, then the framed area is monitored for video recording, but the recorded video is not stored. Not storing the recorded video can prevent invalid video data from occupying storage space.

Example 2

Embodiment 2 is substantially the same as Embodiment 1, the difference is that in Embodiment 2, when the frame selection operation is performed in step 2, the self-service device monitoring area is selected from the monitoring video image of the self-service device, and the business window Surveillance video images respectively frame the outlet staff working area and customer monitoring area.

And, after identifying the face information in step 3, it also includes: comparing the recognized face information of the person with the face information of the branch staff to determine whether the person is a branch employee, if the person is not a branch employee, then consider This person is a customer. In this way, it is avoided to store video when branch employees enter the self-service equipment area or customer monitoring area, and realize efficient recording of the monitoring situation of business windows and self-service equipment.

Based on the results of step 3, in step 4:

The above has schematically described the invention and its implementation. The description is not restrictive, and the invention can be realized in other specific forms without departing from the spirit or basic features of the invention. What is shown in the drawings is only one of the embodiments of the present invention, the actual structure is not limited thereto, and any reference signs in the claims shall not limit the involved claims. Therefore, if a person of ordinary skill in the art is inspired by it, without departing from the purpose of the invention, without creatively designing a structure and an embodiment similar to the technical solution, it shall fall within the scope of protection of this patent. Furthermore, the word "comprising" does not exclude other elements or steps, and the word "a" preceding an element does not exclude the inclusion of "a plurality" of such elements. Multiple elements stated in a product claim may also be realized by one element through software or hardware. The words first, second, etc. are used to denote names and do not imply any particular order.

Claims

A method for monitoring a business window and self-service equipment, comprising the following steps:

Step 1: Record the monitoring video images of the branch business window and the self-service equipment area through the network camera, as the input source for the subsequent steps;

Step 2: Obtain the monitoring video image collected in step 1, and select the area range that needs to be monitored in the monitoring video image;

Step 3: Detect whether there is anyone in the framed area, if there is, identify the person's face information, establish a corresponding identity for the person based on the face information, and save the face information and the corresponding identity;

Step 4: If a person is detected in the framed area in step 3, record the surveillance video of the framed area, and save the recorded video as a group, which is stored based on the identity of the person.
The monitoring method of a kind of business window and self-service equipment as claimed in claim 1, is characterized in that, if in step 3, detects that there is no one in the frame selection area, then in step 4, monitor the frame selection area for video recording .
A kind of monitoring method of business window and self-service equipment as claimed in claim 1, it is characterized in that, preset 3 parameters: detection cycle, alarm cycle and consecutive hit times, in step 3, with detection cycle as time interval, Periodically detect whether there are people in the frame selection area. If there is someone, it will be considered as a hit. In an alarm cycle, if the total number of hits reaches the number of consecutive hits, it will be considered that there are people in the frame selection area.
The monitoring method of a kind of business window and self-service equipment as claimed in claim 1, it is characterized in that, in step 3, select the self-service equipment monitoring area to the monitoring video image frame of the self-service equipment, and the monitoring video image of the business window Select the working area of the branch staff and the customer monitoring area respectively.
A kind of monitoring method of business window and self-service equipment as claimed in claim 1, is characterized in that, step 3 also comprises:

Compare the person's face information with the saved face information, if the person's face information is the same as a saved face information, set the person's identity as the same saved face information The saved identity corresponding to the saved face information.
The monitoring method of a business window and self-service equipment as claimed in claim 1, wherein if no face information is recognized in step 3, then identify the person's portrait features, and establish a The corresponding identification, and save the portrait features and the corresponding identification.
A monitoring method for a business window and self-service equipment as claimed in claim 6, wherein identifying the features of the portrait comprises: the specific method for identifying the features of the portrait is: identifying hair color, hair length, clothing color, clothing pattern, One or more of height and body type.
A kind of monitoring method of business window and self-service equipment as claimed in claim 6, is characterized in that, step 3 also comprises:

Compare the person's portrait features with the saved portrait features, if the person's portrait features are the same as a saved portrait feature, then set the person's identity as the saved face information with the same portrait features The corresponding saved identity.
The monitoring method of a kind of business window and self-service equipment as claimed in claim 4, it is characterized in that, in step 3, if detects that there is someone in the frame area, the face information of the identified person will be compared with the person of the staff of the outlet Face information comparison to determine whether the person is an employee of the branch, if the person is not an employee of the branch, it is considered that the person is a customer;

If it is detected in step 3 that there are people in the monitoring area of the self-service equipment, record the monitoring video in the monitoring area of the self-service equipment, and save the recorded video as a group, which is stored based on the identity of the person;

If it is detected in step 3 that there is no one in the monitoring area of the self-service equipment, record the monitoring video of the monitoring area of the self-service equipment.

If it is detected that there is someone in the customer monitoring area in step 3, and the person is identified as a customer, then the monitoring video of the customer monitoring area is recorded, and the recorded video is saved as a group, which is stored based on the identity of the person ;

If it is detected in step 3 that there are people in the work area of the branch personnel, and the person is identified as a branch employee, then the monitoring video recording is performed on the work area of the branch personnel.
A monitoring device for a business window and self-service equipment, characterized in that it is used to implement the monitoring method for a business window and self-service equipment as claimed in any one of claims 1-9, comprising:

The input module uses the monitoring video images entered in the business window of the network point and the self-service equipment area as the input source for the subsequent steps;

The frame selection module is used to frame the area range that needs to be monitored in the surveillance video image;

The detection module is used to detect whether there is a person in the frame selection area, and if there is a person, identify the face information of the person, establish a corresponding identity mark for the person according to the face information, and save the face information and the corresponding identity mark;

The recording and saving module is used to record the monitoring video of the framed area if a person is detected in the framed area in step 3, and save the recorded video as a group, which is stored based on the identity of the person.