Binocular face recognition intelligent AI camera

Abstract

The invention discloses a binocular face recognition intelligent AI camera which comprises a front shell, a rear cover and a camera module arranged between the front cover and the rear cover, the camera module comprises a support, a first PCBA board and a second PCBA board, the first PCBA board and the second PCBA board are arranged on the support, the first PCBA board is provided with an infrared sensor and an image sensor, the second PCBA board is provided with an SOC chip, and the SOC chip is connected with the front shell. The system comprises an SOC chip and a UVC interface connected with the SOC chip, an ISP processing unit, a DSP processing unit and a storage unit are arranged in the SOC chip, the storage unit is electrically connected with the ISP processing unit and the DSP processing unit, an SDK is embedded in the storage unit, and the infrared sensor and the image sensor are electrically connected with the ISP processing unit of the SOC chip. According to the binocular face recognition intelligent AI camera, an ISP processing unit is integrated in a system chip, the client application cost is reduced, an algorithm is embedded, the computing power of a main control board end does not need to be consumed, the cost of the main control board is reduced, and the cost performance of the chip is improved.

Description

technical field [0001] The invention relates to the technical field of cameras, in particular to a binocular face recognition intelligent AI camera. Background technique [0002] In the prior art, after the camera obtains the face information, the face recognition algorithm needs to perform live detection, face tracking, recognition, comparison and other actions through a PC host plug-in or SDK. This kind of camera is very dependent on the algorithm based on the host. Such a mode has certain hidden dangers, that is, the algorithm security is not high, and it is heavily dependent on the host, and it also has higher requirements on the host CPU, resulting in higher PC costs. In addition, the function and performance of the camera hardware cannot be fully utilized. The camera only handles picture and video communication, with simple functions, low integration and limited security. SUMMARY OF THE INVENTION [0003] The purpose of the present invention is to provide a binocul...

AI Technical Summary

Problems solved by technology

This mode has certain hidden dangers, that is, the algorithm security is not high, and it is heavily dependent on the host, and it also has higher requirements on the host CPU, resulting in higher PC costs.

Moreover, the function and performance of the camera hardware cannot be fully utilized. The camera only processes pictures and video communications, with simple functions, low integration, and limited security.

Images