Click the blue words above to follow us~
The Internet of Things has entered an era of large-scale applications, and more and more industry application scenarios need to achieve cross-hardware architecture development, deployment, and operation of edge computing to achieve higher agility and flexibility. VisionMate has launched an edge computing box solution mainly targeting embedded edge devices that are constrained by size, weight, power consumption requirements, and cost but need to improve performance to support AI workloads, extending computing power to the “last mile” and accelerating the implementation of edge computing scenarios.
VisionMate’s edge computing box solution adopts industrial-grade standard design, has a delicate appearance, compact size, and is sturdy and reliable, suitable for various indoor and outdoor application environments. The lightweight networking solution integrates multiple technologies such as audio and video decoding, data transmission, storage, and intelligent algorithms, with good compatibility, allowing connection to mainstream vendors’ ordinary network cameras, making it easy for users to utilize existing camera resources for AI applications, making monitoring smarter, allowing both local preview and remote control access, with rich interfaces and flexible docking, facilitating third-party system integration.
This solution connects to ordinary network cameras to input video, performs intelligent analysis on objects appearing in the image, outputs abnormal alarm signals, and realizes abnormal behavior alarms:
Supports capturing and comparing faces appearing in the video.
Supports detection of not wearing safety helmets, not wearing work clothes, falls, illegal intrusions, smoking detection, phone calls, crowd gathering, abnormal wandering, and leaving the post.
Supports structured attribute extraction and recognition analysis of vehicles, non-motor vehicles, and items within the scene.
VisionMate can provide edge computing boxes + application SDK to meet the needs of customers with development capabilities to call interfaces for complete application development.
Based on MLU220-M.2, the AIBox-3399DL supports video encoding and decoding, meeting the needs for multi-channel high-definition video (network) encoding and decoding, while having over 8T of computing power, meeting the requirements for multi-channel video image recognition, and supporting more than 4 channels of 1080P video image analysis. It is suitable as an AI processing node to handle multiple videos and reduce the processing pressure on the server. It supports a highly diverse range of artificial intelligence applications such as vision, voice, natural language processing, and traditional machine learning, realizing intelligent edge solutions for various businesses. It provides a complete software framework that can complete functions such as streaming media processing, model conversion, model inference, and model post-processing, and provides relevant examples, paired with high-performance algorithms for target detection, behavior recognition, face recognition, and object recognition, facilitating rapid model deployment and solution development for developers with different needs.
AIBOX-NV21 adopts the NVIDIA Jetson Xavier NX processor, which can provide up to 14 tera operations (TOPS) at a speed of 10W per second or 21 tera operations at a speed of 15W, running multiple neural networks in parallel while processing data from multiple high-resolution sensors. Xavier NX runs on CUDA-X AI software, easily optimizing deep learning networks for edge inference. In terms of CPU, Xavier NX is supported by a 6-core Carmel Arm 64-bit CPU, with 6MB of L2 and 4MB of L3 cache. The device can support up to six CSI cameras on 12 MIPI CSI-2 channels. It comes with 8GB of 128-bit LPDDR4x RAM, capable of data transfer at a speed of 51.2GB / second. The device runs on an Ubuntu-based Linux operating system and provides a complete software framework that can complete functions such as streaming media processing, model conversion, model inference, and model post-processing, and provides relevant examples, paired with high-performance algorithms for target detection, behavior recognition, face recognition, and object recognition, facilitating rapid model deployment and solution development for developers with different needs.
This solution can be widely applied in smart retail, smart parks, smart communities, and other security fields, smart transportation, smart finance, smart grids, smart agriculture, medical devices, small commercial robots, smart cameras, smart factories, and other high-performance AI systems.
VisionMate is a technology company with full-stack capabilities in “algorithms + IoT hardware + software + cloud platform”, dedicated to applying artificial intelligence technology in various practical scenarios, becoming a connector between the physical and digital worlds.
Since its establishment in 2008, after more than ten years of development and accumulation, VisionMate has become a provider of intelligent urban scene digitalization overall solutions. In line with the industry’s requirements for “display + AI”, VisionMate has built a full-stack commercial display and IoT ecosystem platform with hardware, software, applications, and services. Its core business includes the R&D, production, and sales of self-service terminals, and provides AIoT digital full-stack solutions for industry users, including smart commercial displays, smart stores, smart payments, smart campuses, smart communities, smart buildings, smart logistics, and other scenarios.
Corporate Vision: To become a connector between the physical and digital worlds, building an AIoT ecosystem.
Corporate Mission: To make machines better serve humanity and to connect all things intelligently.
Corporate Spirit/Values: Integration, Innovation, Dedication, Success.
Service Purpose: Fast, Timely.
Quality Goal: Professional, Extreme.
Business Philosophy: Empowerment, Symbiosis, Collaboration, Altruism.
Service Areas:
Official Website:
http://www.smdt.com.cn/
Service Hotline:
0755-66890088
Cooperation Consultation
Deputy: 13823335504
Official Email:
scluo@smdt.com.cn
Click the original text for more information!
