Person Detection Using an Ultra Low-resolution Thermal Imager on a Low-cost MCU

TL;DR

This paper presents a lightweight CNN-based person detection method using a low-resolution thermal imager on inexpensive microcontrollers, achieving high accuracy and real-time performance.

Contribution

It introduces a novel ultra-lightweight CNN model trained on thermal data, enabling effective person detection on low-cost hardware.

Findings

Achieves up to 91.62% accuracy (F1-score)

Model has less than 10,000 parameters

Runs in under 100ms on microcontrollers

Abstract

Detecting persons in images or video with neural networks is a well-studied subject in literature. However, such works usually assume the availability of a camera of decent resolution and a high-performance processor or GPU to run the detection algorithm, which significantly increases the cost of a complete detection system. However, many applications require low-cost solutions, composed of cheap sensors and simple microcontrollers. In this paper, we demonstrate that even on such hardware we are not condemned to simple classic image processing techniques. We propose a novel ultra-lightweight CNN-based person detector that processes thermal video from a low-cost 32x24 pixel static imager. Trained and compressed on our own recorded dataset, our model achieves up to 91.62% accuracy (F1-score), has less than 10k parameters, and runs as fast as 87ms and 46ms on low-cost microcontrollers STM32F407 and STM32F746, respectively.