Design Win

Human presence detection using CNN-based computer vision application running on STM32

Solution Description

Designed to determine if a person is present in a given area, human presence detection technology is very useful for a wide range of applications spanning user interfaces, security and safety. It can also be used to actively manage energy consumption by switching on/off lights when someone enters/exits a room or turn on the screen when you approach your thermostat control panel.

Today, there is an increased demand for smart home and smart building solutions to help reduce energy costs as well as provide security and identify opportunities to improve services. Human presence detection applications are a reliable and affordable way to trigger a pre-determined action for an enhanced user experience such as identifying an empty workspace in an office or public library, which is especially useful for social distancing in today's post-pandemic world.

In addition to ensuring a degree of accuracy that will enable immediate action or trusted security alerts, the challenge is to simplify software development and hardware design, especially for battery-based devices with processing, memory, and power constraints. Our solution addresses cost and design considerations by leveraging Artificial Intelligence on cost-effective, ultra-low-power STM32 microcontrollers.

How it works

The system captures digital video from a low-resolution image sensor and ensures efficient processing by a high-performance, dual-core STM32H7 microcontroller running an edge-based human presence detection application (FP-AI-VISION1) based on outputs generated by deep neural networks (DNN) optimized for STM32.

The MCU then displays if a person is detected in the specified area along with the captured video via a connected LCD screen or over a network. The required footprint is very small and allows to add multiple additional features triggered by person detection on the same microcontroller, including but not limited to, people detection and counting, actuation via peripherals, and transmission of information via multiple connectivity options.

The result is a cost-effective and low-power solution for detecting human presence based on an object detection algorithm using a convolutional neural network (CNN) model topology optimized for STM32.

  • Key Product Benefits

    STM32H747XIH6 - High-performance, dual-core, microcontroller with DSP and DP-FPU instructions

    This microcontroller with Arm® Cortex®-M7 and Cortex®-M4 cores has the necessary peripherals to manage incoming digital video signals and the processing power and memory to ensure rapid and accurate people counting applications with minimal power consumption.

    FP-AI-VISION1 - STM32Cube function pack for Computer Vision artificial intelligence (AI) applications 

    The software implements an advanced computer vision application using STM32 neural network libraries based on pre-trained models generated using the STM32CubeMX AI expansion pack (X-CUBE-AI).

  • All Features

    • Ready-to-use firmware featuring the implementation of based on an object detection algorithm using a convolutional neural network (CNN) model for real-time human presence detection applications
    • The edge processing approach ensures lower power consumption and latencies than centralized cloud solutions, with a focus on accuracy and personal privacy
    • Ultra-low-power implementation based on the use of a real-time operating system (RTOS)
    • Easy portability across different STM32 MCU series thanks to the STM32Cube initiative