Talking about the public rental house face recognition of those things

First, how to achieve "live and live"?

The most basic of public rental housing is to protect the safety of the inhabitants. To avoid some security risks, subletting, group renting, and other violations, many public rental housing is not a form of random access. At the same time, in order to reduce the trouble of manual registration operations, many public rental houses have installed face recognition access control systems.

Under the social background of safeguarding and improving people’s livelihood, how to achieve “living and sheltering” has become a problem that the party and the government have tackled and the focus of the broad masses. However, after the construction of public rental housing in our country, there have been many problems in personnel management. .

Question one: The District Housing and Welfare Office cannot effectively manage the people who have already been allocated public rental housing.

Question 2: The Municipal Housing and Social Security Administration needs to make specific plans according to the real-time conditions of public rental housing in various districts.

Second, public rental housing has realized the warning of strangers, lent loans for early warning, vacant rental housing has been alerted

The face recognition system can realize multiple functions such as strangers' warning in the community, early warning of subletting and lending, and vacancy warning after the tenant check-in. The intelligent solution is a kind of biometric technology that identifies people based on their facial features. The camera or camera captures an image or video stream containing faces and automatically detects and tracks the faces in the images, and then performs a series of related techniques on the detected faces, including face image collection, face positioning, and the like. Face feature extraction, face clustering, and face matching achieve the purpose of identifying different identities. Borui's face recognition system employs deep learning technology, hundreds of layers of neural network training, and big data training for thousands of face images. It is a non-contact, non-matching, friendly and dynamic face recognition. system.

Afterwards, the face recognition machine at the entrance of public rental housing can analyze the identity of the commuter by “brushing his face.” If the frequency of strangers reaches a certain value, it indicates that public rental housing may be subleted or loaned; if the tenant does not have a long time “Brushing the face” means that his public rental house is likely to be vacant, and the system will issue an early warning to let staff know the actual living conditions of public rental housing and avoid the waste of public resources.

Third, the classification of public rental tenants and strangers to achieve management

3.1 Personnel Classification Management

In the face recognition system, different name face databases are created, and public rental housing can establish: different face stores such as renters and strangers. When the person enters the door, the camera detects the face image and completes the recognition and displays the belonging library without using cooperation. This scheme can quickly distinguish the personnel information and optimize the personnel management.

3.1.1 Tenant Management

When the tenant rents a house, he goes through the relevant formalities and snaps his face into the warehouse. When he enters or exits the entrance door, he automatically recognizes and records the relevant data. He can set the number of times the tenant exits the gate less than a certain number of times. Managers are suspicious of the status of individual tenants at any time.

3.1.2 The management of strangers

When strangers from outside enter the community, they will automatically capture faces to create a stranger database. When the same stranger enters the community several times, the system will automatically count the number of incoming and outgoing times. Security personnel can manage security measures.

3.2 Personnel Access Management

When a person enters or exits the door, the camera captures an image of the face and generates a recognition result. Based on the recognition result, a daily in and out record is generated. This non-cooperative entry and exit management method facilitates the judgment of the identities of visitors and intelligent management.

3.3 Mobile APP message notification

When the tenant does not appear in the community for a long time or the stranger enters the community more than a set number of times, it can be reminded in the mobile phone APP that the management personnel can receive the notification in the first time and make corresponding preventive measures.

3.4 Display and Linkage Sound

In the intelligent management of public rental housing, different libraries are displayed in different ways and different voices are linked. Through different colors and different sounds with human face display, more clear and eye-catching prompting functions can be achieved.

Fourth, the full application of public rental housing era has come

4.1 System Framework

Face recognition system overall architecture includes front-end equipment, transmission network and monitoring center in three parts, as shown below:

Front-end equipment: It mainly collects face data.

Transmission network: The collected face data is transmitted back to the center.

Monitoring Center: Deploy each service and compare the returned face data in real time.

4.2 System Features

Ø Face database

To achieve the deployment of the staff of the dispatching and control staff, logical division and comparison policy settings can be made according to the attributes of the staff. It also provides users with the ability to create a temporary face database. Users can register themselves and import photos of faces.

Ø Face detection

Automatically detect the face in the video and track and judge the quality, improve the system's recognition performance, and reduce the system's data. You can also analyze existing video files to detect and identify faces. The face detection comparison requires only the use of an ordinary webcam and does not require a special high-definition camera. Supports the camera to capture the passed face photo in real time and retrieve it afterwards. In addition to this, it is also possible to set the camera to capture site photos on a regular basis and provide detailed analysis of the scene, including the characteristics of the suspect's coat and hat, to help confirm the identity of the suspect.

Ø Face recognition

According to the comparison strategy, the face in the captured video is compared with the dispatched list, and if it is compared, the operator is alerted and the operator is prompted to perform the disposition.

Ø Real-time face monitoring

Provides real-time monitoring screens and displays all alarm results. Save all photos of faces that pass through the camera. Each photo is only a few K to a dozen K size.

Ø Capture alarm disposal

The personnel information captured by the camera is further screened and confirmed. Can also record the police's respective operation records, to facilitate the second viewing and confirmation of other personnel.

Ø Face image retrieval

According to the uploaded person image and the search condition, the face database in the corresponding option is searched for a person's face that meets the conditions, and the necessary search conditions such as name, gender, age, and certificate number are input. Query within the conditions can be performed. Supports real-time capture of face photos through the camera and reverse search afterwards.

Ø Portrait information modification

For each different face database, query the information of people who meet the conditions, and modify and delete the information, and register the new personnel information for the selected face database.

Ø System Management Platform

All of the above functional software platforms are integrated and can be installed on a common PC.

Ø Face tracking algorithm

The process of determining the trajectory and size of a face in a video or image sequence.