Edge Vision AI

Advancing Future Safety with Edge Vision AI

With the rapid rise of the AIoT era, artificial intelligence is driving cross-industry integration, expanding the application of IoT technologies across various sectors. The demand for diverse solutions has spurred the rapid evolution of digitization, AI-driven intelligence, and edge-cloud computing technologies. Against this backdrop, AMobile introduces its new Machine Learning series, focusing on Edge Vision AI and setting a new milestone for future intelligent security solutions.

 

AMobile has adopted MediaTek's IoT development platform powered by the Genio 510 chipset, equipped with advanced edge AI processing technologies, including DLA (Deep Learning Accelerator) and VPU (Vision Processing Unit). This technology accelerates AI core computing, significantly enhancing performance and delivering fast and accurate multimedia processing results. The platform is highly flexible, supports various applications, and fully meets the development needs of AI-driven security systems.

Bi-Directional Vision AI for Smart Security

Our bi-directional interactive Vision AI system is designed for smart home security. It incorporates features like facial recognition, de-warping technology, AI noise reduction, and liveness detection, enabling multifaceted data analysis to provide comprehensive home protection.

AI Night Mode

Traditional surveillance cameras often struggle to identify visitors in low-light conditions when unusual outdoor noises occur. We have equipped our system with Sony starlight-grade lenses paired with infrared detection to address this. With Vision AI automatic lighting correction, visitors can be identified even in the dark, eliminating the need for guesswork.

Machine Learning Data for Daytime Analysis

During the day, the system leverages Machine Learning data to recognize visitors' gender, approximate age, and whether they are adults or children. If an unregistered face is detected, the system will start a countdown, and if it remains unresolved after 60 seconds, a remote alert will be triggered. This ensures users can promptly respond to potential threats, preventing intrusions.

AI Liveness Detection

With AI-based liveness detection, the system can accurately distinguish between a natural person and an image used to deceive the camera, protecting against unauthorized entry and potential financial or data theft. This technology effectively prevents identity fraud, offering an additional layer of security.

Dewarp Technology

The system utilizes fisheye lenses without the common problem of dark corners to provide a broader view of outdoor environments. With AMobile's de-warping technology combined with multi-sensor data fusion, Edge Vision AI can output flat images, maintaining high-quality visuals while significantly improving panoramic video analysis and intelligent operational efficiency.

Energy-Efficient and Comprehensive Protection

The AMobile access control system features an IP66-rated waterproof and dustproof design, along with IK08 impact resistance, protecting against impacts from objects weighing up to 1.7 kg within a 0.3-meter range. Additionally, the system supports PoE (Power over Ethernet), eliminating the need for extra power setups, reducing labor costs, and aligning with ESG environmental sustainability standards while ensuring long-term stable operation.

 

We invite you to join AMobile in exploring the future of AI-powered machine vision technology, delving into cutting-edge advancements in AI image analysis systems and intelligent security applications.

Go Back