| Customization: | Available |
|---|---|
| Type: | Professional Software |
| Language Version: | English |
Suppliers with verified business licenses
Audited Supplier Audited by an independent third-party inspection agency
Processor configuration
Each edge AI high-performance processor supports processing 30 video analysis channels or 50 video decoding channels, and can be customized according to the number of front-end video access.
AI training cycle
The system needs to be continuously improved to improve AI recognition rate, and image materials need to be continuously collected for AI training. The upgrade cycle is about 3 months/round.
Algorithm optimization requirements
It is necessary to continuously collect on-site images for periodic work such as annotation, expert appraisal, algorithm optimization, and algorithm re signing, in order to continuously improve the AI recognition rate. Each type of identification object needs to provide no less than 2000 identification images.
Network requirements
1) Internet dedicated line requirements: AI high-performance processors push video transcoding to the cloud, and need to configure an Internet dedicated line (with fixed IP), L (bandwidth)=number of concurrent channels (maximum number of videos simultaneously viewed by remote PC) * 4M (4M for each channel of video) for calculation.
2) LAN networking requirements: Gigabit Ethernet. The network cable uses Category 6 UTP, and each interface of the internal network switch and router should have at least gigabit ports.
Requirements for front-end monitoring equipment
To ensure recognition rate and accuracy, there are specific requirements for the images of front-end video monitoring equipment:
1) The captured image quality is ≥ 720P;
2) Pixels of front-end devices ≥ 4 million;
3) Calculation of effective range for front-end camera monitoring: It is recommended to calculate based on the visible light lens focal length, with effective distance L=f (visible light focal length) * 700.
| Hardware technical parameters of edge Al high-performance processor | |
| CPU | 2 * Intel 10 core with a main frequency of 2.3GHz |
| Memory | 64GB |
| Hard disk | SSD250G+HDD1000G |
| GPU | GTX1080TI 11GB |
| Network card | 1000M dual port Ethernet card |
| Power supply | 800W power supply |
| Operating system | Linux |
| Edge AI high-performance processor software performance indicators | |
| Transcoding video access | ≤ 30 channels |
| Analysis of video access | ≤ 30 channels |
| Computing power support | >13TFLOPS Fp32 support |
| Image analysis | 30 images/second/GPU |
| Supports real-time video analysis | 1 channel/GPU |
| Supports concurrent frame extraction (near real-time) analysis | with 15 channels/GPU |
| Supports timed capture analysis | 30 channels/GPU |
| Capture interval | ≤ 10 seconds |
| Algorithm package | customizable for different analysis objects |
| AI service support | continuously optimizing algorithm packages and customizable extensions |
| Communication protocols | HTTPS, TCP, UDP, etc |
| Authorization method | Hardware encryption dog authorization |
| Support encoding resolution | Dynamic adjustment within the range of 720P-1080P |
| Algorithm framework | Built in AI algorithm frameworks such as Pytorch, TensorFlow, and Darknet |
Chongqing Yingka Electronic Co., Ltd., established in 2007, is located in the Erlang Overseas Student Pioneer Park, Jiulongpo District, Chongqing, China. The company is dedicated to the research and development of wireless sensor networks and AI technology, equipment, and systems for various applications in forestry, scenic areas, and power sectors. These applications include forest fire prevention, forest pest control, wildlife monitoring, bird monitoring, and ecological environment monitoring. Yinka Electronics has successfully overcome key challenges such as long-term field power supply, wireless networking in the field, and reliable low-power sensing. The company has undertaken the "National IoT Application Demonstration Project" by the National Development and Reform Commission and has accumulated over 10 years of experience in forest fire prevention and environmental monitoring applications.
Yinka Electronics is a provider of IoT application solutions for forest fire prevention and environmental monitoring in outdoor environments. It has provided applications and technical services to over 100 scenic areas, forestry, and power units across China.
1. 16 years of experience in the production of IoT sensing devices
2. Experts in the IoT ecosystem industry
3. Collaborate with universities to establish a research team for expert academician workstations
4. Professional business personnel can respond online at any time
5. Comprehensive after-sales service
6. Customizable development and fast delivery
Online guidance and video tutorials for overseas buyers, with 24-hour online response
Q1: Are you manufacturer?
A: (yes)
Q2: What is the delivery cycle?
A: Usually one month
Q3: How do you ship the goods and how long does it take arrive?
A: We usually ship by DHL, UPS, FedEx or by Air. 3-5 days after payment received.
Q4: Can I have a sample order?
A:Yes,we welcome sample order to test and check quality.
Q5: Do you offer OEM service?
A: Yes, we can design the things according to your request.
){kind=link}