R&D Factory for Face Recognition Edge Computing Box

Edge computing boxes are different fromtraditional face recognition machines. Traditional face recognition machinesare often used for attendance checking or access control, and they do notrequire computing power. Most of them perform static recognition and comparisonof pre-recorded face recognition images. In contrast, face recognition applications running on edge computing boxes with built-in 1T computing poweror higher 6T computing power usually involve facial recognition and peoplecounting. When applied to construction sites, they can perform human shaperecognition, action recognition, safety helmet recognition, and PersonnelStranded Identification, etc.   

In addition, edge computing boxes generallyperform dynamic recognition of camera streams and video streams by connectingto external cameras or camcorders, and conduct analysis through the AI analysisand computing unit of the NPU to improve recognition efficiency. The camerascan be USB cameras or POE IP cameras. Through the RTSP protocol, the analysisand computing of image frames related to AI applications can be obtained.  

The NPUs of RK3568 and RK3588 both suppor the hardware-accelerated AI recognition and deep learning. The aforementioned connected camera devices, POE IP cameras, monitors, etc., can all be regardedas traditional AIoT devices, which together form a complete AIoT project.  

In terms of operating systems, Dingchang's edge computing boxes support Android systems to accelerate AI algorithm andapplication development, as well as Linux Ubuntu systems for the samepurpose.  

Dingchang's edge computing boxes and the internal RK3568 and RK3588 motherboards are all independently developed andmanufactured. The motherboards come in a variety of models with complete functions. In particular, the RK3588 edge computing box model (as shown in thefigure below) features an all-aluminum alloy exterior with excellent heat dissipation, making it one of the best-selling edge computing box models and styles.  

In addition, the editor would like to emphasize that Dingchang provides a complete set of AI algorithm support platforms and APIs for in-depth AI ecological partners. In addition to enablingcustomers to quickly get started with AI algorithm development, deep learning,and large language models on the RK3568 and RK3588 hardware platforms, it also offers dozens of case DEMO codes, including text OCR, text-to-speechrecognition, semantic segmentation, commodity recognition, license platerecognition, face detection, human shape + head recognition, human posture,action recognition, QR code recognition, urban street scene target detection,etc.  

The above is just a small part of thefunctions and technical support of edge computing-related products mentioned bythe editor from Dingchang. In terms of production, cooperative customers canalso rest assured of our production capacity.

We have established large-scale physicalprocessing and manufacturing factories in Guangzhou and Hunan. We welcome newand old customers and leaders at all levels to visit our factories nearby forinspection and guidance. We are dedicated to serving you.