Development of a Baby Surveillance System using Video Processing
Keywords:
Video processing, Baby surveillance system, Optical flowAbstract
This research developed a baby surveillance system in a crib bed aiming to detect normal or abnormal movements of baby especially actions that may cause accidents to young children. The system can automatically interpret gestures in the specified areas, which keep an eye on the safety of the child in bedroom. The system using video processing can analyze video frames to detect children behaviors. For example, if a baby is climbing then the system will alert parents. The method of baby surveillance system consists of taking an input video of a baby to use in the system, zoning specific areas on video and eliminating image noises. The following steps are detecting the movements of the child by using the optical flow and analyzing the movements of baby with flow and direction. The system is implemented on OPENCV library and evaluated with video from single-view stationary camera, which is a single view. For the experiment, we used 10 videos of real events at home from the YouTube website. In the videos, a baby was in the cage bed and had normal movements such as climbs and falls. Each video had different environment. The experimental results showed that the baby monitoring system can analyze normal postures, climbing and falling postures with an accuracy of 80.7%.
References
Cheng, L., Sveta, Z., Tjon W. E., & Peter, H.N. (2016). Video-based Discomfort Detection for Infants Using a Constrained Local Model. International Conference on Systems, Signals and Image Processing (IWSSIP) (pp. 1-4). Bratislava, Slovakia.
Daniel, T., Kjersti, E., Ivar, A., & Oyvind, M. (2016). Motion Based Detection of Respiration Rate in Infants using Video. IEEE International Conference on Image Processing (ICIP) (pp.1225-1229). Phoenix, AZ, USA.
Eren, A., Cagan, C., Elif, A., Huseyin, I., Doga, G., & Duy, I. (2020). ThermoCam: Smart Baby Monitoring Assistant. IEEE 44th Annual Computers, Software, and Applications Conference (COMPSAC) (pp. 636-643). Madrid, Spain.
Gunnar, F. (2003). Two-Frame Motion Estimation Based on Polynomial Expansion. In: Scandinavian Conference on Image Analysis (pp. 363-370). Gothenburg, Sweden.
Leong, L., Hussain, A., Zulkifley, M., & Zaki, W. (2015). Camera-Based Toddler Fall Detection System by using Kalman Filter, Journal of Theoretical and Applied Information Technology, 81(2), 383-388.
Muhammad, M., Syahrull, S., Muhammad, R., & Addzrull, S. (2012). Clinical Infant Pain Trial Based on k-NN Algorithm. International Symposium on Instrumentation & Measurement, Sensor Network and Automation (IMSNA), 2, 371-373.
Natchanan, C. (2016). Accidents in Children: Situation and Prevention. The Journal of Faculty of Nursing Burapha University, 24(3), 1-12.
PETacular.(2019). Babies Escaping Cribs Video Compilation (Video file). Retrieved from https://www.youtube.com/watch?v=5oaxNuOChdY
Rabua, C., Siham, M., Oussama, A., & El hadi, K. (2020). An Intelligent Baby Monitoring System based on Raspberry PI, IoT Sensors and Convolutional Neural Network. 2020 IEEE 21st International Conference on Information Reuse and Integration for Data Science (IRI) (pp. 365-371). Las Vegas, NV, USA.
Statistics of Child Health Emergencies. (2016). Newsletter of the National Institute for Emergency Medicine. Retrieved from http://www.niems.go.th/1/Upload/migrate/File/ 256103221110087879 TOQBStXAB8C3NZcF.pdf
Yang, M., & Chuang M. (2013). Fall Risk Assessment and Early-Warning for Toddler Behaviors at Home. Sensors, 13(12), 16987-17002.
Yogita, D., & Sachin, D. (2019). Baby Monitoring System using Image Processing and IoT. International Journal of Engineering and Advanced Technology (IJEAT), 8(6), 4961-4964.
