Islak Alanlarda Mahremiyet Koruyucu Düşme Tespiti: Düşük Çözünürlüklü Termal Diziler ve Uyarlanabilir Eşikleme Kullanan IoT Entegreli Bir Sistem

Öz

Banyo ve tuvalet gibi ıslak zeminlerde meydana gelen düşme kazaları, özellikle yaşlı ve engelli bireyler için ciddi yaralanmalara ve hatta can kayıplarına yol açan önemli bir halk sağlığı sorunudur. Bu alanlarda mahremiyet kaygıları nedeniyle geleneksel optik kameraların kullanımı mümkün olmamaktadır. Bu çalışmada, mahremiyeti ihlal etmeden banyo ortamındaki düşme eylemlerini tespit eden, Nesnelerin İnterneti entegreli düşük maliyetli bir sistem geliştirilmiştir. Geliştirilen sistem; Raspberry Pi uç birimi, düşük çözünürlüklü termal kamera modülü ve sesli uyarı bileşenlerinden oluşmaktadır. Yazılım mimarisinde ise Python tabanlı görüntü işleme algoritmaları ve kullanıcı etkileşimi için Flutter tabanlı bir mobil uygulama kullanılmıştır. Önerilen algoritma, bireyin vücut yüzey alanı ve ısı haritası yoğunluğunu analiz eder; zemindeki sıcak piksel yoğunluğunun belirlenen kritik eşik değerini (%40) aşması durumunda eylemi "düşme" olarak sınıflandırır. Tespit edilen acil durumlar, MQTT protokolü aracılığıyla anlık olarak mobil uygulamaya iletilerek bakıcı veya sağlık ekiplerinin en kısa sürede müdahale etmesi sağlanır. Deneysel sonuçlar, sistemin ayakta durma, eğilme ve düşme eylemlerini başarıyla ayırt edebildiğini ve mahremiyet koruyucu bir can yeleği işlevi görebileceğini doğrulamaktadır.

Anahtar Kelimeler

• Islak Zemin
• IoT
• Termal Görüntüleme
• Görüntü İşleme

Privacy-Preserving Fall Detection in Wet Areas: An IoT-Integrated System Using Low-Resolution Thermal Arrays and Adaptive Thresholding

Öz

Falls occurring on wet surfaces such as bathrooms and toilets are a significant public health issue, particularly for elderly and disabled individuals, leading to serious injuries and even fatalities. The use of traditional optical cameras in these areas is not feasible due to privacy concerns. In this study, a low-cost system integrated with the Internet of Things (IoT) has been developed that detects falls in the bathroom environment without violating privacy. The developed system consists of a Raspberry Pi edge device, a low-resolution thermal camera module, and an audio alert component. The software architecture utilizes Python-based image processing algorithms and a Flutter-based mobile application for user interaction. The proposed algorithm analyzes the individual's body surface area and heat map intensity; it classifies the action as a “fall” when the density of hot pixels on the floor exceeds a specified critical threshold value (40%). Detected emergencies are instantly communicated to the mobile application via the MQTT protocol, ensuring that caregivers or medical teams can intervene as quickly as possible. Experimental results confirm that the system can successfully distinguish between standing, bending, and falling actions and can function as a privacy-preserving life jacket.

Anahtar Kelimeler

• Wet Floor
• IoT
• Thermal Imaging
• Image Processing

Kaynakça

1. [1] Türkiye İstatistik Kurumu, “Sağlık Harcamaları İstatistikleri, 2021,” TÜİK Kurumsal, Ankara, Türkiye, 2021. [Çevrimiçi]. Erişim: https://data.tuik.gov.tr/Bulten/Index?p=SaglikHarcamalari-Istatistikleri-2021-45728
2. [2] Z. Haytoğlu, “Ev kazaları ve korunma,” Çukurova Medical Journal, c. 46, s. 1–3, 2021.
3. [3] A. Sarohan, Acil servise ev kazası ile gelen hastalarda dikkat eksikliği ve hiperaktivite bozukluğu ilişkisinin değerlendirilmesi, Yüksek Lisans Tezi, Türkiye, 2017.
4. [4] S. Shilaskar, S. Bhatlawande, R. Chavare, R. Joshi ve S. Jaiswal, “Vision based fall detection with BRISK feature descriptor,” 2023 Second International Conference on Electrical, Electronics, Information and Communication Technologies (ICEEICT), ss. 1– 7, 2023.
5. [5] G. Cao, J. Yang ve M. Chen, “Fusion system design of multi-sensors and visual algorithm for elderly fall Şekil 15. Uygulamada başka bir düşme tespiti çıktısıVeri Bilimi Derg, 9(1), 1-11, 2026 11 detection,” 2023 IEEE 6th Eurasian Conference on Educational Innovation (ECEI), ss. 146–148, 2023.
6. [6] B. Wang ve Y. Guo, “Soft fall detection using frequency modulated continuous wave radar and regional power burst curve,” 2022 Asia-Pacific Microwave Conference (APMC), ss. 240–242, 2022.
7. [7] F. A. Elshwemy, R. Elbasiony ve M. T. Saidahmed, “A new approach for thermal vision based fall detection using residual autoencoder,” International Journal of Intelligent Engineering & Systems, c. 13, s. 2, 2020.
8. [8] L. Yu, H. Chen, H. He, H. Nie, X. Zhai ve B. Xiong, “A fall detection system based on a thermopile imaging array and a back projection algorithm,” 2020 IEEE International Conference on Electro Information Technology (EIT), ss. 60–65, 2020.

9. [9] P. Lynggaard, “A fall-detection system that uses body area network and thermal energy harvesting technologies,” 2018 11th CMI International Conference, ss. 67–73, 2018.
10. [10] Y. Taniguchi, H. Nakajima, N. Tsuchiya, J. Tanaka, F. Aita ve Y. Hata, “A falling detection system with plural thermal array sensors,” 2014 Joint 7th International Conference on Soft Computing and Intelligent Systems (SCIS) and 15th International Symposium on Advanced Intelligent Systems (ISIS), ss. 673–678, 2014.
11. [11] C. Taramasco vd., “A novel monitoring system for fall detection in older people,” IEEE Access, c. 6, ss. 43563–43574, 2018.
12. [12] F. Riquelme, C. Espinoza, T. Rodenas, J. G. Minonzio ve C. Taramasco, “eHomeSeniors dataset: An infrared thermal sensor dataset for automatic fall detection research,” Sensors, c. 19, s. 20, 4565, 2019.
13. [13] A. M. Rezaei vd., “An unobtrusive fall detection system using low resolution thermal sensors and convolutional neural networks,” 2021 43rd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), ss. 6949–6952, 2021.
14. [14] E. Vermeulen, Automatic Incident Detection (AID) with Thermal Cameras, 2014.
15. [15] A. R. Putri vd., “Compressed neural network for thermal array-based fall detection system on embedded AI,” 2021 International Conference on Information and Communication Technology Convergence (ICTC), ss. 1754–1757, 2021.
16. [16] J. Adolf, M. Macas, L. Lhotska ve J. Dolezal, “Deep neural network based body posture recognitions and fall detection from low resolution infrared array sensor,” 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), ss. 2394– 2399, 2018.
17. [17] X. Luo, T. Liu, J. Liu, X. Guo ve G. Wang, “Design and implementation of a distributed fall detection system based on wireless sensor networks,” EURASIP Journal on Wireless Communications and Networking, ss. 1–13, 2012.
18. [18] J. Nogas, S. S. Khan ve A. Mihailidis, “Fall detection from thermal camera using convolutional LSTM autoencoder,” IJCAI Workshop on Aging, Rehabilitation and Independent Assisted Living, 2018.
19. [19] S. Kido, T. Miyasaka, T. Tanaka, T. Shimizu ve T. Saga, “Fall detection in toilet rooms using thermal imaging sensors,” 2009 IEEE/SICE International Symposium on System Integration (SII), ss. 83–88, 2009.
20. [20] V. Mehta, A. Dhall, S. Pal ve S. S. Khan, “Motion and region aware adversarial learning for fall detection with thermal imaging,” 2020 25th International Conference on Pattern Recognition (ICPR), ss. 6321–6328, 2021.
21. [21] S. Denkovski vd., “Multi visual modality fall detection dataset,” IEEE Access, c. 10, ss. 106422– 106435, 2022.
22. [22] C. Zhong vd., “Multi-occupancy fall detection using non-invasive thermal vision sensor,” IEEE Sensors Journal, c. 21, s. 4, ss. 5377–5388, 2020.
23. [23] A. Hayashida, V. Moshnyaga ve K. Hashimoto, “New approach for indoor fall detection by infrared thermal array sensor,” 2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS), ss. 1410–1413, 2017.
24. [24] A. Naser, A. Lotfi, M. D. Mwanje ve J. Zhong, “Privacypreserving, thermal vision with human in the loop fall detection alert system,” IEEE Transactions on Human-Machine Systems, c. 53, s. 1, ss. 164–175, 2022.
25. [25] N. Bharathiraja vd., “Real-time fall detection using ESP32 and AMG8833 thermal sensor: A nonwearable approach for enhanced safety,” 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS), ss. 1732–1736, 2023.
26. [26] A. Hayashida, V. Moshnyaga ve K. Hashimoto, “The use of thermal IR array sensor for indoor fall detection,” 2017 IEEE International Conference on Systems, Man, and Cybernetics (SMC), ss. 594–599, 2017.
27. [27] J. Rafferty vd., “Thermal vision based fall detection via logical and data driven processes,” 2019 IEEE International Conference on Big Data, Cloud Computing, Data Science & Engineering (BCD), ss. 35–40, 2019.
28. [28] B. D. Reading ve B. Freeman, “Simple formula for the surface area of the body and a simple model for anthropometry,” Clinical Anatomy, c. 18, s. 2, ss. 126–130, 2005.