STMicroelectronics introduces STM32 support for computer vision rapid development tools

     STMicroelectronics introduces a new AI firmware feature pack and camera module hardware suite to enable embedded developers to create affordable and powerful computer vision applications that run on edge devices based on STM32 * microcontrollers (MCUS).

     The STM32Cube feature pack FP-AI-VISION1 contains several complete computer vision application code examples that run convolutional neural networks (CNNS) on the STM32H747 and are easily portable across the full STM32 family of products. The firmware presents several application examples where developers can retrain neural networks with selected data sets, providing greater freedom and flexibility to solve a variety of use case problems.

     New features include support for USB VC cameras (webcam mode) to simplify image acquisition tasks, as well as food classification and user presence detection code samples, where user presence detection examples create convenient visual "wake-up calls" to wake the system from power-saving mode. The STM32 Wiki has an article on how to create an image classification application using the Teachable Machine online tool in combination with the STM32Cube.AI and FP-AI-VISION1 feature packs.

     The B-CAMS-OMV camera kit works best with FP-AI-VISION1 firmware and provides the hardware needed to train and deploy neural network models. The camera kit includes a transfer card with a built-in ST MB1379 5-megapixel OV5640 color camera module. The adapter is compatible with all STM32 Discovery and evaluation boards with ZIF interfaces, and can also be used with the VG5661 vehicle gauge grayscale global shutter camera from ST. In addition, the Waveshare interface and OpenMV interface allow users to connect a variety of third-party infrared and visible light cameras to solve a wider range of computer vision applications. There is an article on how to integrate STM32Cube.AI generated code into the OpenMV ecosystem on the STM32 Wiki.

     FP-AI-VISION1 includes a variety of frame buffer processing capabilities, camera drivers, as well as image capture software, pre-processing software, and neural network inference software, and several neural network models are available, including floating-point based models and X-CUBE-AI generated quantitative models. X-CUBE-AI is ST's optimized artificial neural network C code generator, which allows developers to fine-tune neural models for intended applications because of its support for flexible memory configurations.