International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 11 | Nov 2024
p-ISSN: 2395-0072
www.irjet.net
REAL TIME HUMAN ACTIVITY RECOGNITION FOR ELDERLY CARE IN SMART HOME Maria Sobana.S1, Krishna Veni.M2, Mirsha Shelyn.J3, Nagarani.V.O4 1Assistant Professor, Department of Computer Science and Engineering,
K.L.N. College of Engineering, Sivagangai, India
2,3,4Student, Department of Computer Science and Engineering,
K.L.N. College of Engineering, Sivagangai, India ---------------------------------------------------------------------***--------------------------------------------------------------------Abstract - Recognizing and predicting human activities in smart homes is increasingly important for enhancing resident’s quality of life and safety. One key feature of smart home systems is the ability to recognize human activities in real-time, enabling automation, improving energy efficiency, and offering personalized assistance. This model introduces a method for real-time human activity recognition using the K-Nearest Neighbors (K-NN) algorithm in a smart home environment. Sensor data from various sources, such as motion, pressure, temperature and sound sensors is collected to identify and classify different activities performed by residents. The K-NN algorithm is chosen for its simplicity, robustness, and effectiveness in handling multi-dimensional data. By analyzing patterns in sensor data, the system accurately predicts activities such as walking, sitting, cooking and falling. The process includes sensor data collection, data preprocessing, feature extraction, and the implementation of K-NN for classification. The systems performance is evaluated using a dataset collected from a simulated smart home, demonstrating high accuracy in real-time activity recognition. This approach is particularly useful for applications like elderly care where understanding and predicting human behavior is critical for ensuring safety and efficiency.
1.1 MOTION SENSOR The motion sensor detects any physical movement within a defined area. This is crucial for tracking whether an individual is active or stationary, enabling the system to monitor daily activities such as walking, entering or exiting a room, or detecting prolonged inactivity. By identifying these movements, the motion sensor helps in distinguishing between regular behaviors and potential concerns. These sensors are typically placed in high-traffic areas like hallways, living rooms, or entryways to capture comprehensive movement data across the home. 1.2 PRESSURE SENSOR A pressure sensor measures the force or weight applied to a surface, making it especially useful for determining whether a person is sitting, standing, or lying down. For example, it can detect when someone sits on a chair or lies on a bed, providing valuable information for recognizing regular activities, such as rest or relaxation, and abnormal events like sudden collapses. These sensors are often placed on chairs, beds, or on the floor in key locations where individuals frequently sit or lie down. This helps the system track important behavioral patterns and detect irregularities in posture.
Key Words: Real-Time Human Activity Recognition, Machine Learning, Sensors, KNN Algorithm, Elderly Care.
1.3 TEMPERATURE SENSOR The temperature sensor continuously monitors the ambient temperature in various areas of the home. This is useful for detecting environmental changes that may indicate specific activities, such as an increase in temperature when someone is cooking or using an appliance, or a sudden drop when a window is left open or if the heating system fails. In elderly care, this sensor can also detect changes in room temperature that may impact the individual's comfort or health, such as extreme cold or heat, which could signal a dangerous situation. These sensors are typically placed in living areas, kitchens, or other rooms where temperature changes are critical to monitor.
1. INTRODUCTION Smart home technologies play a crucial role in enhancing the quality of life for elderly individuals living alone, addressing the growing need for remote monitoring solutions due to the challenges faced by youngsters who live far from their elderly parents. In this context, the proposed system leverages the K-Nearest Neighbors (KNN) algorithm for real-time Human Activity Recognition (HAR) in smart homes. By analyzing sensor data, the system effectively classifies a range of activities, including walking, sitting, sudden falls etc. The primary objective is to ensure the safety of elderly individuals while providing peace of mind to their family members through real-time updates on their elderly people activities.
© 2024, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 32