As the foundation of smart city construction, the Internet of Things is a prerequisite for smart perception in future cities, but the huge data it brings is beyond the reach of human analysis. If the data cannot be transformed into meaningful information, it is meaningless. AI technology just plays an important role in processing data. Artificial intelligence can correctly identify, find anomalies, and predict the future. Through artificial intelligence combined with the non-siwan of the gate hardware, it can realize the integration of artificial intelligence and industrial applications and realize the development of applied intelligence.
Face application function:
1) Real-time capture. Based on the front-end high-definition camera or face capture camera, the system or capture camera detects the face in real-time video, tracks the movement of the face, and captures the clearest frame for storage. And record the captured face photos, elapsed time, camera location information, etc. in the passerby library, and the captured and stored face information can be used as a search database. Support to select the capture channel according to the tree-shaped target, and view one or more real-time facial image captures at the same time. Support the downloading of background images and small images
. 2) Real-time warning (face bayonet). Support real-time comparison of captured pictures and blacklist library. Support the setting of early warning reception, in the early warning setting, you can choose the deployment control task and deployment control range of the early warning reception.
3) Historical warning. Supports single condition or combined condition query according to deployment control task, deployment scope, deployment control object, similarity, time, and alarm confirmation form. Support to set the query results to be sorted by time or similarity.
4) Face query. Supports face query for dynamic snapshot library and static list library. Query photos support original image view, detailed information view, and video preview before and after. Face images and related structured information can be exported into excel files.
5) Search face by face (1:N comparison). Users can select a certain portrait picture and search for portrait pictures with high similarity in the snapshot library or static list library. The system sorts according to the similarity. The pictures to be compared can be uploaded locally, or they can be snapped pictures or static pictures. When the uploaded image is too blurry, support the user to manually mark the feature to strengthen the recognition function, manually mark the feature points or frame selection range through the website interface, help the system to recognize the accurate face position, improve the comparison accuracy rate, and improve the comparison of blurred photos Effect.
6) Face duplicate check (N:N comparison). The system supports a single personnel database or repeated personnel queries between two personnel databases, and returns the duplicate check results. In the process of the duplication check task, you can view the task status, related information, etc., and perform operations such as viewing and deleting the completed duplication check task.
7) Face APP. Support the face retrieval function, and compare faces by uploading photos or local pictures. After the comparison is successful, the corresponding face retrieval results will be returned according to the similarity.
8) Personnel trajectory analysis. You can use the existing face pictures or the face pictures retrieved by the system to search for similar face pictures within a certain period of time and monitoring range, select the target person’s face picture, and analyze the target person’s “from where and where to go” , Where to pass along the way”. (Fei Si Wan)