IR-UWB Sensor Based Fall Detection Method Using CNN Algorithm
Abstract
:1. Introduction
- High performance of detection algorithm: accuracy, sensitivity, specificity, F1 score, etc.
- User convenience.
- Privacy preservation.
1.1. Sensors
1.2. Algorithms
1.3. Characteristics of IR-UWB Radar Sensor and Related Works
1.4. Our Approaches
- Proposed a preprocessing method that converts one-dimensional time series signal data of IR-UWB radar sensor to a distance-time two-dimensional image, which contains features of signal change effectively.
- -
- Decision of standard frame of pixel figure difference, decision of motion image window size, proposition of labeling method.
- Proposed to use generated 2D grey scale image by CNN algorithm for fall detection.
- -
- Extraction and learning of signal change pattern of radar sensor by CNN algorithm automatically.
- Proposed a decision algorithm that can decide the type of event by reviewing the classification result of a single image.
- -
- Replaced problem of image classification for problem of fall detection and proposed the image judgement ratio that shows peak performance of an experiment.
2. Materials and Methods
2.1. System Overview
2.2. Stage 1: Motion and Data Acquisitions
2.3. Stage 2: Data Pre-Processing
- Perform inter-frame subtraction after 24th frame;
- Determine size of motion image;
- Outlier replacement and normalizing;
- Grey scale image visualization and image data labelling for CNN algorithm.
2.4. Stage 3: CNN Algorithm Based Feature Extraction and Classification
2.4.1. CNN Architecture for Training and Classification
2.4.2. Performance Evaluation Levels: Single Image and Single Event
2.4.3. Performance Evaluation Indicators
3. Experimental Setup
4. Results and Discussion
4.1. Motion Image Visualization
4.2. Database Configuration for CNN-Based Model Construction and Evaluation
4.2.1. Image-Level Database Configuration
4.2.2. Event-Level Database Configuration
4.3. Classification Results
4.3.1. Classification Results for Single Image
4.3.2. Classification Results for Single Event: Evaluating the Final Performance of Model
5. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- World Health Organization. Global Report on Falls Prevention in Older Age; World Health Organization Press: Geneva, Switzerland, 2008; Available online: http://www.who.int/ageing/projects/Falls_prevention7March.pdf (accessed on 20 October 2020).
- Mubashir, M.; Shao, L.; Seed, L. A survey on fall detection: Principles and approaches. Neurocomputing 2013, 100, 144–152. [Google Scholar] [CrossRef]
- Xu, T.; Zhou, Y. Elders’ fall detection based on biomechanical features using depth camera. Int. J. Wavelets Multiresolution Inf. Process. 2018, 16, 1840005. [Google Scholar] [CrossRef]
- Igual, R.; Medrano, C.; Plaza, I. Challenges, issues and trends in fall detection systems. BioMed. Eng. Online 2013, 12, 66. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Lord, C.J.; Colvin, D.P. Falls in the elderly: Detection and assessment. Proc. Ann. Conf. Eng. Med. Biol. 1991, 13, 1938–1939. [Google Scholar]
- Seredin, O.S.; Kopylov, A.V.; Huang, S.C.; Rodionov, D.S. A skeleton features-based fall detection using microsoft kinect V2 with one class-classifier outlier removal. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2019, XLII-2/W12, 189–195. [Google Scholar] [CrossRef] [Green Version]
- Bourke, A.K.; O’Brien, J.V.; Lyons, G.M. Evaluation of a threshold-based tri-axial accelerometer fall detection algorithm. Gait Posture 2007, 26, 194–199. [Google Scholar] [CrossRef]
- Lai, C.F.; Chang, S.Y.; Chao, H.C.; Huang, Y.M. Detection of cognitive injured body region using multiple triaxial accelerometers for elderly falling. IEEE Sens. J. 2011, 11, 763–770. [Google Scholar] [CrossRef]
- Bianchi, F.; Redmond, S.J.; Narayanan, M.R.; Cerutti, S.; Lovell, N.H. Barometric pressure and triaxial accelerometry-based falls event detection. IEEE Trans. Neural Syst. Rehab. Eng. 2010, 18, 619–627. [Google Scholar] [CrossRef] [PubMed]
- Li, Q.; Zhou, G.; Stankovic, J.A. Accurate, fast fall detection using posture and context information. In Proceedings of the 6th ACM Conference on Embedded Networked Sensor Systems—SenSys’08, Raleigh, NC, USA, 5–7 November 2008; pp. 443–444. [Google Scholar]
- Nyan, M.N.; Tay, F.E.; Murugasu, E. A wearable system for pre-impact fall detection. J. Biomech. 2008, 41, 3475–3481. [Google Scholar] [CrossRef]
- Noury, N.; Fleury, A.; Rumeau, P.; Bourke, A.K.; Laighin, G.Ó.; Rialle, V.; Lundy, J.E. Fall detection—Principles and methods. In Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology, Lyon, France, 22–26 August 2007; pp. 1663–1666. [Google Scholar]
- Alarifi, A.; Ayed Alwadain, A. Killer heuristic optimized convolution neural network-based fall detection with wearable IoT senosor devices. Measurement 2021, 167, 108258. [Google Scholar] [CrossRef]
- Hashim, H.; Mohammed, S.; Gharghan, S. Accurate fall detection for patients with Parkinson’s disease based on a data event algorithm and wireless sensor nodes. Measurement 2020, 156, 107573. [Google Scholar] [CrossRef]
- Ren, L.; Peng, Y. Research of fall detection and fall prevention technologies: A systematic review. IEEE Access 2019, 7, 77702–77722. [Google Scholar] [CrossRef]
- Sposaro, F.; Tyson, G. iFall: An android application for fall monitoring and response. In Proceedings of the 31st Annual International Conference of the IEEE Engineering in Medicine and Biology Society: Engineering the Future of Biomedicine, Minneapolis, MN, USA, 2–6 September 2009; pp. 6119–6122. [Google Scholar]
- Casilari-Pérez, E.; García-Lagos, F. A comprehensive study on the use of artificial neural networks in wearable fall detection systems. Expert Syst. App. 2019, 138, 112811. [Google Scholar] [CrossRef]
- Espinosa, R.; Ponce, H.; Gutiérrez, S.; Martínez-Villaseñor, L.; Brieva, J.; Moya-Albor, E. A vision-based approach for fall detection using multiple cameras and convolutional neural networks: A case study using the UP-Fall detection dataset. Comput. Biol. Med. 2019, 115, 103520. [Google Scholar] [CrossRef] [PubMed]
- Rougier, C.; Meunier, J.; St-Arnaud, A.; Rousseau, J. Fall detection from human shape and motion history using video surveillance. In Proceedings of the 21st International Conference on Advanced Information Networking and Applications Workshops—AINAW’07, Niagara Falls, ON, Canada, 21–23 May 2007; pp. 875–880. [Google Scholar]
- Chua, J.L.; Chang, Y.C.; Lim, W.K. A simple vision-based fall detection technique for indoor video surveillance. Signal Image Video Process. 2015, 9, 623–633. [Google Scholar] [CrossRef]
- Feng, G.; Mai, J.; Ban, Z.; Guo, X.; Wang, G. Floor Pressure Imaging for Fall Detection with Fiber-Optic Sensors. IEEE Pervasive Comput. 2016, 15, 40–47. [Google Scholar] [CrossRef]
- Droghini, D.; Ferretti, D.; Principi, E.; Squartini, S.; Piazza, F. A combined one-class SVM and template-matching approach for user-aided human fall detection by means of floor acoustic features. Comput. Intell. Neurosci. 2017, 2017. [Google Scholar] [CrossRef] [Green Version]
- Wang, H.; Zhang, D.; Wang, Y.; Ma, J.; Wang, Y.; Li, S. RT-Fall: A Real-Time and Contactless Fall Detection System with Commodity WiFi Devices. IEEE Trans. Mob. Comput. 2016, 16, 511–526. [Google Scholar] [CrossRef]
- Wild, K.; Boise, L.; Lundell, J.; Foucek, A. Unobtrusive in-home monitoring of cognitive and physical health: Reactions and perceptions of older adults. J. Appl. Gerontol. 2008, 27, 181–200. [Google Scholar] [CrossRef]
- Lindemann, U.; Hock, A.; Stuber, M.; Keck, W.; Becker, C. Evaluation of a fall detector based on accelerometers: A pilot study. Med. Biol. Eng. Comput. 2005, 43, 548–551. [Google Scholar] [CrossRef]
- Kangas, M.; Konttila, A.; Lindgren, P.; Winblad, I.; Jämsä, T. Comparison of low-complexity fall detection algorithms for body attached accelerometers. Gait Posture 2008, 28, 285–291. [Google Scholar] [CrossRef] [PubMed]
- Mrozek, D.; Koczur, A.; Malysiak-Mrozek, B. Fall Detection in Older Adults with Mobile IoT Devices and Machine Learning in the Cloud and on the Edge. Inf. Sci. 2020, 537, 132–147. [Google Scholar] [CrossRef]
- Doukas, C.; Maglogiannis, I.; Tragas, P.; Liapis, D.; Yovanof, G. Patient fall detection using support Vector Machines. In Proceedings of the IFIP International Conference on Artificial Intelligence Applications and Innovations, Athens, Greece, 19–21 September 2007; pp. 147–156. [Google Scholar]
- Albert, M.; Kording, K.; Herrmann, M.; Jayaraman, A. Fall classification by machine learning using mobile phones. PLoS ONE 2012, 7, e36556. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Hussain, F.; Umair, M.B.; Ehatisham-ul-Haq, M.; Pires, I.M.; Valente, T.; Garcia, N.M.; Pombo, N. An Efficient Machine Learning-based Elderly Fall Detection Algorithm. arXiv 2019, arXiv:1911.11976. [Google Scholar]
- Cheng, J.; Xiang, C.; Chen, X.; Shen, M. A Framework for Daily Activity Monitoring and Fall Detection Based on Surface Electromyography and Accelerometer Signals. IEEE J. Biomed. Health Inform. 2012, 17, 38–45. [Google Scholar] [CrossRef] [PubMed]
- Wang, X.; Gao, C.; Guo, Y. Elderly fall detection using SIFT hybrid features. In AOPC 2015: Image Processing and Analysis; International Society for Optics and Photonics: Bellingham, WA, USA, 2015; p. 96751W. [Google Scholar] [CrossRef]
- Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet classification with deep convolutional neural networks. Commun. ACM 2017, 60, 84–90. [Google Scholar] [CrossRef]
- Ma, L.; Liu, M.; Wang, N.; Wang, L.; Yang, Y.; Wang, H. Room-level fall detection based on ultra-wideband (UWB) monostatic radar and convolutional long short-term memory (LSTM). Sensors 2020, 20, 1105. [Google Scholar] [CrossRef] [Green Version]
- Zhang, T.; Zhang, Q.; Xu, H.; Zhang, H. A practical ranging method using IR-UWB signals. In Proceedings of the 2013 9th International Wireless Communications and Mobile Computing Conference—IWCMC 2013, Sardinia, Italy, 1–5 July 2013; pp. 1839–1844. [Google Scholar]
- Choi, J.W.; Cho, S.H. A new multi-human detection algorithm using an IR-UWB radar system. In Proceedings of the 2013 3rd International Conference on Innovative Computing Technology—INTECH 2013, London, UK, 29–31 August 2013; pp. 467–472. [Google Scholar]
- Li, Z.; Li, W.; Lv, H.; Zhang, Y.; Jing, X.; Wang, J. A novel method for respiration-like clutter cancellation in life detection by dual-frequency IR-UWB radar. IEEE Trans. Microw. Theory Tech. 2013, 61, 2086–2092. [Google Scholar] [CrossRef]
- Hu, X.; Jin, T. Short-range vital signs sensing based on EEMD and CWT using IR-UWB radar. Sensors 2016, 16, 2025. [Google Scholar] [CrossRef] [Green Version]
- Silva, B.; Hancke, G.P. IR-UWB-based non-line-of-sight identification in harsh environments: Principles and challenges. IEEE Trans. Ind. Inform. 2016, 12, 1188–1195. [Google Scholar] [CrossRef]
- Leem, S.K.; Khan, F.; Cho, S.H. Vital sign monitoring and mobile phone usage detection using IR-UWB radar for intended use in car crash prevention. Sensors 2017, 17, 1240. [Google Scholar] [CrossRef] [Green Version]
- Khan, F.; Leem, S.; Cho, S.H. Hand-based gesture recognition for vehicular applications using IR-UWB radar. Sensors 2017, 17, 833. [Google Scholar] [CrossRef]
- Ahmed, S.; Khan, F.; Ghaffar, A.; Hussain, F.; Cho, S.H. Finger-counting-based gesture recognition within cars using impulse radar with convolutional neural network. Sensors 2019, 19, 1429. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Ota, K.; Ota, Y.; Otsu, M.; Kajiwara, A. Elderly-care motion sensor using UWB-IR. In Proceedings of the 2011 IEEE Sensors Applications Symposium, San Antonio, TX, USA, 22–24 February 2011; pp. 159–162. [Google Scholar]
- Mercuri, M.; Soh, P.J.; Pandey, G.; Karsmakers, P.; Vandenbosch, G.A.; Leroux, P.; Schreurs, D. Analysis of an indoor biomedical radar-based system for health monitoring. IEEE Trans. Microw. Theory Tech. 2013, 61, 2061–2068. [Google Scholar] [CrossRef] [Green Version]
- Mokhtari, G.; Aminikhanghahi, S.; Zhang, Q.; Cook, D.J. Fall detection in smart home environments using UWB sensors and unsupervised change detection. J. Reliab. Intell. Environ. 2018, 4, 131–139. [Google Scholar] [CrossRef]
- Tsuchiyama, K.; Kajiwara, A. Accident detection and health-monitoring UWB sensor in toilet. In Proceedings of the 2019 IEEE Topical Conference on Wireless Sensors and Sensor Networks—WiSNet, Orlando, FL, USA, 20–23 January 2019; pp. 1–4. [Google Scholar]
- Wang, Y.; Wu, K.; Ni, L.M. Wifall: Device-free fall detection by wireless networks. IEEE Trans. Mobile Comput. 2016, 16, 581–594. [Google Scholar] [CrossRef]
- LeCun, Y.; Bottou, L.; Bengio, Y.; Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 1998, 86, 2278–2324. [Google Scholar] [CrossRef] [Green Version]
- Boutellaa, E.; Kerdjidj, O.; Ghanem, K. Covariance matrix based fall detection from multiple wearable sensors. J. Biomed. Inform. 2019, 94, 103189. [Google Scholar] [CrossRef]
- Chen, L.; Li, R.; Zhang, H.; Tian, L.; Chen, N. Intelligent fall detection method based on accelerometer data from a wrist-worn smart watch. Measurement 2019, 140, 215–226. [Google Scholar] [CrossRef]
- Islam, M.M.; Tayan, O.; Islam, M.R.; Islam, M.S.; Nooruddin, S.; Kabir, M.N.; Islam, M.R. Deep Learning Based Systems Developed for Fall Detection: A Review. IEEE Access 2020, 8, 166117–166137. [Google Scholar] [CrossRef]
- Hakim, A.; Huq, M.S.; Shanta, S.; Ibrahim, B.S.K.K. Smartphone based data mining for fall detection: Analysis and design. Proced. Comput. Sci. 2017, 105, 46–51. [Google Scholar] [CrossRef]
Type of Sensor | Advantages | Disadvantages |
---|---|---|
Wearable Devices |
|
|
Camera-based sensor |
|
|
Ambient Sensor |
|
|
Class | Behavior Type | Definition |
---|---|---|
ADL | Walk | Move at a regular pace by lifting and setting down each foot in turn, never having both feet off the ground at once |
Sit | The act of placing one’s body on the floor by weighting one’s hips in straight posture (one’s back is upright) | |
Rise | The act of standing up straight after a “fall” or “sit” | |
Stand | The act of staying straight with one’s legs stretched out against the ground for a while | |
Fall | Fall | Move downward, typically rapidly and freely without control, from higher (standing state) to a lower level |
Indicator | Description | Calculation |
---|---|---|
True positive (TP) | A fall occurs, the system detects it | Count |
True negative (TN) | A normal (ADL) movement is performed, the system does not declare a fall | Count |
False negative (FN) | A fall occurs but the system does not detect it | Count |
False positive (FP) | The system announces a fall, but it did not occur | Count |
Accuracy (A) | Ratio of correctly predicted observations to total observations | |
Recall (R) | Ratio of the positive observations correctly predicted to all observations in the actual class | |
Precision (P) | Ratio of correctly predicted positive observations of the total predicted positive observations | |
F1 score (F1) | Weighted average of precision and recall |
Participant No. | Gender | Age | Height [cm] | Weight [kg] |
---|---|---|---|---|
1 | Female | 22 | 158 | 48 |
2 | Female | 22 | 161 | 60 |
3 | Female | 23 | 167 | 54 |
4 | Male | 21 | 165 | 63 |
5 | Male | 23 | 173 | 72 |
6 | Male | 24 | 175 | 69 |
Parameter | Value | |
---|---|---|
Sensor Module | Detecting range | Up to 13 m |
Frequency range | 3 ~ 4 GHz | |
Bandwidth | 0.45 ~ 1 GHz | |
Output power | Typ. −25 dBM | |
Distance resolution | 2.03 cm | |
Dimension | 17.5.5 mm × 63.82 mm × 10 mm | |
Antenna Specification | Type | UWB directional antenna |
Gain | Avg. 7 dBi | |
Antenna angle | 56° (X-Z plane), 77.5° (Y-Z plane) | |
Size | 76 mm × 58.5 mm × 10 mm |
Hyper Parameter | Description |
---|---|
Convolution filter size | 5 × 5 |
Stride | 1 |
Epochs | 100 |
Batch size | 50 |
Dropout rate | 10% |
Type of Dataset | Class | Motion Type | Quantity | Ratio | |
---|---|---|---|---|---|
Total Dataset (447,360) | Train dataset (339,288, 75.84%) | Fall (68,280, 20.12%) | Fall | 68,280 | 20.12% |
ADL (271,008, 79.88%) | Walk | 114,144 | 33.64% | ||
Stand | 21,360 | 6.30% | |||
Sit | 45,360 | 13.37% | |||
Rise | 90,144 | 26.57% | |||
Test dataset (108,072, 24.16%) | Fall (22,176, 20.36%) | Fall | 22,176 | 20.52% | |
ADL (85,896, 79.48%) | Walk | 34,992 | 32.38% | ||
Stand | 7440 | 6.88% | |||
Sit | 14,568 | 13.48% | |||
Rise | 28,896 | 26.74% |
Class | Quantity | Ratio | |
---|---|---|---|
Total events generated (1560) | Fall | 282 | 18.08% |
ADL | 1278 | 81.92% |
Predicted | |||
---|---|---|---|
Positive | Negative | ||
Observed | Positive | (True Positive) 16436 | (False Negative) 5740 |
Negative | (False Positive) 5441 | (True Negative) 80,455 |
Indicators | Value |
---|---|
Classification accuracy for training data | 99.99% |
Accuracy(A) | 89.65% |
Recall(R) | 74.12% |
Precision(P) | 75.13% |
F1 score (F1) | 74.62% |
Predicted | |||
---|---|---|---|
Positive | Negative | ||
Observed | Positive | (True Positive) 260 | (False Negative) 22 |
Negative | (False Positive) 35 | (True Negative) 1243 |
Indicators | Value |
---|---|
Accuracy(A) | 96.65% |
Recall(R) | 92.20% |
Precision(P) | 88.14% |
F1 score (F1) | 90.12% |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Han, T.; Kang, W.; Choi, G. IR-UWB Sensor Based Fall Detection Method Using CNN Algorithm. Sensors 2020, 20, 5948. https://doi.org/10.3390/s20205948
Han T, Kang W, Choi G. IR-UWB Sensor Based Fall Detection Method Using CNN Algorithm. Sensors. 2020; 20(20):5948. https://doi.org/10.3390/s20205948
Chicago/Turabian StyleHan, Taekjin, Wonho Kang, and Gyunghyun Choi. 2020. "IR-UWB Sensor Based Fall Detection Method Using CNN Algorithm" Sensors 20, no. 20: 5948. https://doi.org/10.3390/s20205948