Thermo presence

Detect and count people on infrared images from a low-resolution thermal camera (24x32 pixels).

This is the official repo for the paper: Low-Cost Thermal Camera-Based Counting Occupancy Meter Facilitating Energy Saving in Smart Buildings

Overview

Using passive infrared sensors is a well-established technique of presence monitoring. While it can significantly reduce energy consumption, more savings can be made when utilising more modern sensor solutions coupled with machine learning algorithms. This paper proposes an improved method of presence monitoring, which can accurately derive the number of people in the area supervised with a low-cost and low-energy thermal imaging sensor. The method utilises U-Net-like convolutional neural network architecture and has a low parameter count, and therefore can be used in embedded scenarios. Instead of providing simple, binary information, it learns to estimate the occupancy density function with the person count and approximate location, allowing the system to become considerably more flexible. The tests show that the method compares favourably to the state of the art solutions, achieving significantly better results.

Table of contents

• Dataset
• Data collection
• Data processing
• Hardware evaluation
• Citation

Dataset

Our dataset is publicly available in the dataset directory in the form of HDF files which consist of data arrays (thermal images) and labels (people positions' coordinates). The raw thermal data and illustrative RGB images are available in the cloud shared folder. Moreover, the dataset/README.md includes training, validation and test sequences split and example data loading using Python. The table below shows the summary of training, validation, and test datasets considering the number of people in the frame.

	0	1	2	3	4	5	Total
Training	99	105	2984	3217	1953	114	8472
Validation	0	139	631	1691	225	139	2825
Test	162	83	211	341	1235	315	2347

Data collection

Software to record the data from a camera. To be deployed on the Raspberry Pi device. See data_collection/README.md

Data processing

Software to process/analyse the data. See data_processing/README.md

Hardware evaluation

Script to evaluate a model on Thermo Presence dataset. Tested on Raspberry Pi 4B with Intel Neural Compute Stick 2 and Google Coral USB Accelerator. See evaluation/README.md

Citation

@Article{en14154542,
    AUTHOR = {Kraft, Marek and Aszkowski, Przemysław and Pieczyński, Dominik and Fularz, Michał},
    TITLE = {Low-Cost Thermal Camera-Based Counting Occupancy Meter Facilitating Energy Saving in Smart Buildings},
    JOURNAL = {Energies},
    VOLUME = {14},
    YEAR = {2021},
    NUMBER = {15},
    ARTICLE-NUMBER = {4542},
    URL = {https://www.mdpi.com/1996-1073/14/15/4542},
    ISSN = {1996-1073},
    DOI = {10.3390/en14154542}
}