Raspberry Pi, known for its affordable single-board computers, has unveiled a new add-on that enhances vision-based AI applications: the Raspberry Pi AI Camera. Priced at $70, this innovative image sensor integrates on-board AI processing capabilities.
The AI Camera features a Sony image sensor (IMX500) and is powered by the RP2040 microcontroller, Raspberry Pi’s efficient and cost-effective chip. While it won’t replace high-end Nvidia GPUs for complex AI tasks, this module enables users to capture and process images using popular neural network models, making it suitable for various applications.
One significant advantage of the on-board processing is that it allows the host Raspberry Pi to remain available for other tasks, eliminating the need for a separate accelerator. The new AI Camera is compatible with all Raspberry Pi models.
This isn’t Raspberry Pi’s first foray into camera technology; the company continues to offer the Raspberry Pi Camera Module 3, which features a simple 12-megapixel Sony image sensor (IMX708) and retails for around $25. The AI Camera shares the same dimensions (25mm x 24mm) as its predecessor but is slightly thicker due to the optical sensor design. It comes pre-loaded with the MobileNet-SSD model for real-time object detection.
Originally targeted at tech enthusiasts and hobbyists, Raspberry Pi now primarily serves companies that incorporate its devices into their products or utilize them in industrial settings. The new AI Camera expands the possibilities for vision-based applications across various sectors.
