In addition to the hard standards for the accuracy of face recognition and the storage of face database, the equipment manufacturers participating in the bidding are also required to provide a software platform to support the information supervision of public rental housing.

At the end of 2018, the face recognition system of public rental housing in Beijing was officially put into operation, and some public rental housing residents began to “brush their faces” to enter and exit the community and unit building doors. According to Beijing affordable housing construction investment center, Beijing has installed face recognition systems for 66 public rental housing projects. In 2019, face recognition systems will be installed for nearly 200 public rental housing projects in the city.

Beijing vigorously promotes the construction of affordable housing face recognition projects with a clear original intention, mainly to improve the supervision level of community violation early warning, dynamic monitoring, statistical analysis and so on with the help of face recognition technology. It is also to further accelerate the informatization of public rental housing use supervision, improve management methods, improve supervision efficiency, realize effective control over the actual residents of public rental housing through face recognition, and prevent vacancy, sublease and lending.

The application of face recognition in affordable housing such as public rental housing is an effective measure to implement the real person system of affordable housing, which can further ensure the fair use of social resources. At the same time, the launch of the face recognition access control system is a further supplement to the community security construction and improves the overall security of the community and the convenience of residents’ life.

Face recognition is used to assist real person management. At present, it has been widely used in government affairs such as affordable housing, social security and pension. For face recognition that has entered the public’s clothing, food, housing and transportation, we can clearly feel the convenience it brings to us. However, at the same time, you may be unfamiliar with the engineering deployment behind it. In fact, when a seemingly simple face recognition system project is deployed, the details are not simple, because it involves the practicability of the engineering project, so the rules and regulations will have many detailed standards and requirements. Here, we take the public rental housing project in Beijing as an example.

Technical requirements of Beijing public rental housing face recognition system

Beijing’s public rental housing face recognition project has detailed technical requirements in the bidding document. In addition to the hard standards of the system itself in face recognition accuracy and face database storage, the equipment manufacturers participating in the bidding also need to provide a software platform to support public rental housing information supervision.

Taking the public rental housing face recognition system project of Beijing Economic and Technological Development Zone as an example, the bidding procurement mainly includes 360 sets of face recognition cameras, 94 sets of face recognition analyzers and 1 set of public rental housing dynamic supervision system. These equipment will be installed in 11 public rental housing projects on demand. Among them, the face camera is mainly installed on the first floor unit door and underground unit door, serving 62 buildings and 11835 tenant families in total.

1. Recognition accuracy

Recognition accuracy refers to the accuracy obtained by comparing the image information obtained by the camera with the image information in the face database. The recognition accuracy is divided into daytime recognition accuracy and nighttime recognition accuracy. The recognition accuracy is directly related to the accuracy of basic data acquisition, which affects the results of data analysis. The project requires that the recognition accuracy of non face recognition shall not be less than 90% in the daytime and 85% at night.

2. Effective capture rate

Effective capture rate refers to the ratio of the number of effective portraits captured by the camera to the number of people passing through the camera shooting range over a period of time. When the recognition accuracy is high, the higher the effective capture rate, the higher the accuracy of basic data acquisition. The effective capture rate of the project is required to be no less than 90%. This indicator is under the condition that the face is unobstructed and undisturbed, the head lowering angle is no more than 20 °, and the left and right faces are no more than 45 °.

3. Capture range

According to the actual situation of public rental housing project, it is necessary to capture and identify personnel within 5m long and 5m wide.

4. Identity authentication and access control

It is required to realize the unique identity authentication of public rental family personnel, and control the opening of access control by comparing the obtained identity information with the collected and registered base information. At the same time, it has alarm service for abnormal conditions.

The bid inviter also lists detailed technical requirements for related products of back-end face recognition analyzer:

In addition, in terms of supporting public rental housing dynamic supervision system software, the project also lists 15 software application requirements, including visual unified management, query and display, personnel access log management, equipment inspection management, household verification management subsystem, arrears collection information service, rent collection management, automatic generation of public rental housing dynamic supervision report, etc, In order to realize the real-time visual and unified management of the occupancy and rent payment of public rental houses in the region and the early warning of various violations.

From the above procurement requirements and technical requirements of some public rental face recognition systems in Beijing Economic and Technological Development Zone, it can be seen that compared with traditional security surveillance cameras and NVR equipment, users have higher technical standards and functional demands for front and rear end equipment of face recognition. Of course, the technical requirements of face recognition system will vary according to local conditions in different cities and different public rental housing projects. The security function of the system is only the foundation. The core of the system deployment is to implement the real person system of public rental housing through face recognition and realize the dynamic supervision of public rental housing by the housing authority.

As China enters the “accelerated run” stage of affordable housing construction, Beijing, Shenzhen, Hangzhou, Shanghai and other places have vigorously promoted the construction of affordable housing such as public rental housing in recent years. In order to strengthen the management demand for public rental housing in the whole region, intelligent management will be put on the agenda. Among them, the real-time management platform based on face recognition will become the top priority in the next few years, Face recognition will usher in rapid development in the public rental housing market!

Leave a Reply

Your email address will not be published. Required fields are marked *