Application Scenario

Due to its small size, rich expansion functions and high cost performance, CAM-CRV1126S2U/CAM-CRV1109S2U has a wide range of application scenarios. Such as: face recognition gates, smart door locks, gesture 4 recognition, smart travel, etc. The following introduces several mainstream application scenarios of CAM-CRV1126S2U/CAM-CRV1109S2U.

Face Recognition Gate

The application scenarios of face recognition gates are divided into two cases: external screen and bare metal. The two application scenarios are compatible with each other, and there is no need to modify the software and hardware for both connections. Just select and adapt the face recognition algorithm that suits you before use. Below we have adapted the ArcSoft algorithm as an example to introduce the application scenarios of face recognition gates. This chapter only introduces specific algorithm adaptation, please see the “Application Development” section.PS: The officialFacial_GateFirmware corresponds to this scene.

External Screen

The external screen allows users to have a complete and good experience. The screen supports touch, and more operations can be done directly on the screen after a complete interactive interface is built. The following is a real machine display.

  • Adapt to high-quality RGB camera with professional ISP image quality debugging. The following is the recognition preview effect.

_images/pass.jpg

  • The adaptation of the IR camera effectively prevents cheating on photos, videos and wax figures.

_images/npass.jpg

  • Typec otg provides hardware support for background control, face registration management, and gate background maintenance.

_images/weblogin.png

Bare Metal Gate

Bare metal refers to no external screen. All functions of the bare metal machine are the same except for user interaction through the screen. It is even possible to modify the debugging serial port to a normal serial port to achieve the function of face recognition and result output without connecting to any device. In addition, the face library can still be imported into the device through Typec otg. This scenario can minimize the cost of face recognition gates.

AI_UVC Camera

CAM-CRV1126S2U/CAM-CRV1109S2U can virtualize itself as a UVC and RNDIS device, through the Typec otg interface, it can be connected to any host computer to make a AI_UVC camera.The following is the effect under the windows platform.PS: The officialUVCFirmware corresponds to this scene

  • Face detection.

_images/face.jpg

AI Computing Stick

The CAM-CRV1126S2U/CAM-CRV1109S2U motherboard itself can also be used as an AI calculation stick, which can assist the host computer to perform some AI calculations by connecting to the host computer. The device connects to the host through Typec otg as a slave device, and calls npu auxiliary operations through the official RKNN-Toolkit. For details, please refer to the section “RKNN-Toolkit - Quick Start”.