Horus AI Camera

Horus AI Camera

Horus is a set of software and hardware platforms customized for AI cameras. With an RV1126 quad-core processor, these platforms use mature software and hardware designs, have stable quality, and provide cost-effective solutions that help our customers accelerate their productization process and enable rapid product deployment to seize market opportunities.

1. Overview

Smart Camera is a highly integrated micro machine vision system. This device integrates image acquisition, processing and communication functions into a single camera, including video acquisition, data storage, analysis, and data transmission.

The smart camera has a built-in artificial intelligence chip, image acquisition unit, image processing unit, and data transmission unit, which can analyze video content in real time, detect moving objects, and identify attribute information of people and car. The device processes and analyzes the collected image information, and then transmit the structural data to the server for secondary integrated analysis.

Horus Camera Structure

Smart cameras have the following characteristics:

1. Compact structure, high integration, stable performance, low failure rates; computing power can be expanded according to actual usage requirements.

2. As the smart camera is used close to the user side, it can collect information that affects the phenomenon for analysis in time, and report the data information that the user cares about in real time, showing high real-time performance and low network bandwidth occupancy.

3. The smart camera has low cost and easy installation, allowing quick, multi-point installation as well as distributed and fast expansibility.

4. Through the back-end server, smart cameras at different distribution points can be networked, and real-time information interaction can be carried out between the cameras in the network, showing interactive characteristics.

5. The smart camera can remotely set the gain, electronic shutter, and shutter time in the background to achieve precise external synchronization control of the exposure time.

6. The smart camera can be equipped with multiple digital I/O, Ethernet, RS232 ports, USB, and other interfaces to expand the local storage space.

Product System Block Diagram

Horus is a set of software and hardware platform customized for AI Camera. This product has carried out professional board signal test, power supply quality test, environmental adaptability test, and product heat dissipation simulation and temperature rise test during the R&D phase. Customers can use Smart cameras exert their maximum performance. Horus has built-in Linux operating system, detailed product testing, stable overall performance, and complete software and hardware design data, which facilitates customers to quickly complete algorithm transplantation, accelerate product application execution, improve productization speed, and quickly take products to market.

Horus uses the RV1126 main control chip, quad-core A7 processor, 1GB memory, 8GB storage, 2T powerful computing power, with 5 million or 2 million image sensors, which the customers can choose according to their application scenarios. Horus reserves dual sensor interfaces, infrared supplementary light interface, wired network interface supports POE power supply and 4 KV lightning protection, making it safer and more reliable to use.

horus 1
Horus adopts standard 38-board design, and a variety of product appearances can be freely selected

2. Application Scenarios

In recent years, the smart camera industry has developed rapidly and has been widely used in various industries. Horus series products have been widely used in the following application scenarios.

2.1. Passenger Flow Analysis

The smart camera can be used in shopping malls, supermarkets, retail, and tourism venues for the collection of passenger flow information. Smart cameras are installed at the entrances and exits of the business area and designated areas to accurately collect, analyze, store and transmit passenger flow data. In a stable indoor lighting environment, the accuracy of passenger flow statistics can reach more than 95% without post-adjustment. The device can be remotely configured and managed through the web interface, and also supports real-time collection of passenger flow data in eight areas, heat map display, and customer residence time statistics. The collected data, combined with customized report analysis, can provide a basis for the transformation of various entities to new retail.

Facial Feature Recognition
Facial Feature Recognition

2.2. Behavior Analysis

Based on face recognition and body shape recognition, the device enables analyses  such as crowd density analysis and crowd gathering analysis. For example, such analysis in areas or places with a large number of people can prevent crowd incidents, and achieve early warning and timely notification of management personnel. During a epidemic, such analysis can effectively help prevent the gathering of people and spread of the infectious disease.

Body Feature Recognition
Behavior Analysis

2.3 Hard Hat Recognition

All companies take various measures to ensure the personal safety of employees and production safety. Injuries and deaths caused by not wearing a hard hat can occur, and the management of wearing a hard hat has been challenging. To improve the safety awareness of work personnel, hard hat recognition devices can be installed in various production sites to allow early warning from the real-time video that monitor the safety precautions in accordance with the requirements.

horus文档图 4
Hard Hat Recognition