Arducam IMX500 AI Camera Module
SKU
B0642-1
- [MCU-FIRST AI] The first robust solution to bring full IMX500 AI performance to microcontrollers (ESP32-P4, Arduino, etc.), bridging the gap between sensors and SoCs.
- [DUAL INTERFACE] Unique design featuring both SPI and MIPI outputs, offering maximum compatibility across low-end MCUs and high-end processors.
- [INDUSTRIAL STABILITY] A cost-effective, high-stability embedded AI camera solution, optimized for mass production and harsh environments.
- [POWER EFFICIENT] On-sensor processing drastically reduces data transmission load, enabling ultra-low power consumption ideal for battery-operated IoT edge devices.
- [DEVELOPER READY] Features a refined C++ SDK and AI Model zoo support to minimize setup time, replacing complex legacy modules with a streamlined workflow.
Overview
Specifications
Wiki
Documents
Video
Customization
Overview
Arducam, Edge, AI Camera.
Where they intersect, magic happens.
Arducam
Arducam’s latest innovation transforms the Sony IMX500 into a truly universal AI camera series. Breaking free from the limitations of previous experimental modules, this new platform is engineered to bridge the ecological gap between microcontrollers (MCUs) and high-end SoCs, offering a standardized, scalable vision solution.
Edge
Redefining the boundaries of embedded vision. Unlike traditional solutions that rely on heavy operating systems, this module enables MCU-grade AI. It delivers instantaneous on-device processing, ultra-low power consumption, and industrial-grade stability—perfect for scenarios where privacy, latency, and energy efficiency are paramount.
AI Camera
The heart of the module is Sony's IMX500 Intelligent Vision Sensor, integrating pixel sensing and logic computation on a single chip. We have unlocked its full potential for the embedded world, allowing developers to deploy sophisticated AI models on platforms like ESP32 P4, Arduino, and Raspberry Pi without the complexity of a full Linux stack.
Applications
These application examples are for reference only; please contact us for MORE specific details.
Note: Applicable to industrial scenarios that only focus on model inference results
1. Industrial Object Detection
- Smart Metering: Reading analog dials and digital meters in legacy equipment.
- Defect Detection: Identifying anomalies on production conveyor belts (Quality Control).
- Safety Monitoring: Detecting the presence of humans in hazardous robot operating zones.
- Logistics: Counting packages and tracking cargo movement in warehouses.
2. Segmentation & Analysis
- Agricultural IoT: Leaf segmentation for disease detection and crop growth monitoring on low-power edge devices.
- Privacy-First Monitoring: Segmenting people from backgrounds to transmit only activity data without compromising identity.
- Road Condition Analysis: Identifying lane markings and obstacles for autonomous guided vehicles (AGVs).
3. Pose & Gesture Estimation
- Contactless Control: Operating kiosks or industrial panels via hand gestures to maintain hygiene.
- Worker Safety: Monitoring ergonomic posture to prevent workplace injuries.
- Fitness Wearables: analyzing movement form on battery-powered devices.
4. Classification & Sorting
- Waste Management: Classifying trash types for automated sorting bins.
- Inventory Management: Identifying products on shelves for automated stock-taking.
- Predictive Maintenance: Visual classification of machine parts wear and tear.
Features
Arducam’s One-stop Vision AI Solution
Powered by Sony's IMX500 Intelligent Vision Sensor
Optimized for MCU (ESP32, Arduino) & Embedded Platforms
Compatible with SPI and MIPI Interfaces
- 1. SMART
Unleashing AI on MCUs
Historically, running complex AI on microcontrollers was impossible. Not anymore. By offloading neural network processing to the IMX500 sensor itself, this module liberates the MCU's resources. It serves as a high-performance AI Camera for cost-sensitive and power-constrained industrial applications. Supported by Arducam’s Model Zoo, it allows for the deployment of pre-trained models—bridging the gap between lightweight hardware and heavy-duty AI performance. - 2. SHARP
Industrial-Grade Precision
Designed for the rigors of real-world deployment, the module leverages the IMX500’s 12.3MP resolution to capture critical details in varying lighting conditions. Whether used in automated production lines or outdoor monitoring, it ensures consistent data output. While currently in a flexible kit form factor, it is engineered for stability, ensuring reliable inference results for professional integrators. - 3. VERSATILE
The Universal Interface (SPI + MIPI)
True versatility means playing well with others. This module is distinctively designed with both SPI and MIPI interfaces, making it the "Swiss Army Knife" of AI cameras.
For MCUs: Connect via SPI to platforms like ESP32 P4 or Arduino for low-power IoT applications.
For SoCs: Use the MIPI interface for higher bandwidth integration.
Power Flexibility: Supports Type-C power delivery and external dual-mode power, accommodating diverse industrial electrical standards. - 4. USER-FRIENDLY
Simplified Development Experience
We’ve learned from the past. Moving away from the complexity of the previous IMX500 MCU AI Camera series, this new generation features a Cross-Platform C++ SDK. It is designed to lower the learning curve, offering a development experience akin to standard UVC cameras but for embedded systems. With customization support for multiple requirements, developers can move from unboxing to proof-of-concept in record time, regardless of their preferred hardware platform.
Specifications
| Camera | |
|---|---|
| Sensor | IMX500 |
| Resolution | 12.3 Megapixels |
| Chroma | Color |
| Shutter Type | Rolling Shutter |
| Optical Format | 1/2.3" |
| Output Format | RAW/YUV |
| Pixel Size | 1.55µm x1.55µm |
| Sensor Active Area | 4056(H)×3040(V) |
| Sensor Image Area | 6.29 mm x 4.71 mm (7.86 mm diagonal) |
| Input Tensor Size (SPI - AI Metadata) |
Input tensor size Min 64x48; Input tensor size Max 640x480; |
| Resolution@fps on ESP32 P4 (MIPI - IMAGE) |
IMAGE_ONLY_MODE: 1920x1080@30FPS, 1600x1200@30FPS, 1280x720@30FPS; 1024x600@30FPS, 640x480@30FPS; WITH_AI_MODE: 1600x1200@10-20FPS, 1280x720@10-20FPS; 1024x600@10-20FPS, 640x480@10-20FPS |
| Lens | |
| Focus Type | Fixed Focus |
| FOV(DxHxV) | 82°(D)×72°(H)×56°(V) |
| Lens Mount | M12 Mount |
| F.NO | 3.0±5% |
| E.F.L | 4.3±5%mm |
| IR Sensitivity | Integral 650nm IR Filter, visible light only |
| TV-Distortion | <0.58% |
| Functional and Compatibility | |
| Interface | SPI, I2C, MIPI, USB Type-C |
| Supported Platforms | ESP32 P4, etc. |
| SDK Support | Support |
| AI Model Support | Arducam AI Model Zoo |
| Electrical and Mechanical | |
| Camera Board Size | 34mm x 34mm |
| Overall Weight (Net) | 27.3g |
| Overall Size | 34mm x 34mm x 34.2mm |
| Standard and Certifications | |
| Temperature range (Storage) | -4°F-167°F (-20°C to +75°C) |
| Temperature range (Operating) | 32°F-122°F (0°C to +50°C) |
| Certification | RoHS |
Package Including
- 1 x Arducam IMX500 Edge AI Camera Module for MCU&SoC
- 1 x 15cm 15-22pin FFC Cable
- 1 x 10cm 8pin Dupont Cable
Documentation
- Product Datasheet
- Arducam IMX500 Camera Series AI Model Zoo
- Arducam IMX500 SPI AI Camera SDK
- ESP32-P4 Function EV Board User Guide
- ESP32-P4 Function EV Board (Github)
- 3D Drawing
Customization
About Arducam's Customization Services:
If off-the-shelf solutions don't meet your needs, Arducam offers end-to-end custom camera design services to help you turn your ideas into real products. From early concept to mass production, we work closely with you to tailor an imaging solution to meet your unique technical requirements and product goals.
What can Arducam customize?
Sensor/Lens/circuit/interface;
Optical filter/housing/ISP, and more.
For more information about Arducam's custom design, please refer to: Arducam Custom Design