XCORE-VISION Hits the Sweet Spot for Computer Vision Applications
Computer vision projects often force developers to choose between two extremes. Simple microcontroller-based camera boards are inexpensive and power efficient, but they quickly hit a wall when it comes to more advanced image processing. A Raspberry Pi or similar single-board computer offers considerably more processing power, but brings an operating system, longer boot times, and additional power consumption along with it. XMOS is targeting the space between these options with XCORE-VISION.
XCORE-VISION is a small embedded vision platform built around the XMOS XCORE.AI XU316 multicore microcontroller. Powered by this chip, the board can capture camera data and run AI models locally in real time without relying on cloud services.
A closer look at the hardware (📷: XMOS)
The XCORE architecture is well suited to this task because it supports hardware-level parallel processing and deterministic execution. Developers can divide different portions of an application across processing resources and know when operations...
Copyright of this story solely belongs to hackster.io. To see the full text click HERE