Mobile Device Fall Detection via Sensor Fusion and Image Obfuscation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fall detection systems are invasive, inconvenient, and lack accuracy, particularly for individuals who spend most of their time alone at home, due to stigma, complexity, and high costs, and often require manual intervention or wearables that may not be suitable for all users.
Innovation Solution
A system using a mobile electronic device with an energy sensor to capture and analyze images of a space, employing spatial and body measurements to automatically detect falls or other physical movements, while minimizing privacy concerns through image obfuscation and storage, and transmitting alerts without storing images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If wearable fall detection devices are used, then automatic fall detection capability is provided, but user acceptance and adoption rate are low due to stigma and inconvenience
Solution Approach 1:
The patent introduces an intermediary device (mobile electronic device such as smartphone or tablet) that mediates between the user and the fall detection function. Instead of requiring dedicated wearable sensors, the system uses the user's existing mobile device with its built-in sensors to detect falls, thereby reducing stigma while maintaining detection capability
Solution Approach 2:
The system leverages the multi-functionality of mobile electronic devices that users already carry daily. The same device used for communication and entertainment also performs fall detection, eliminating the need for separate specialized wearable devices and improving user acceptance
2Ease of operation
If manual fall detection devices are used, then user control is maintained, but detection accuracy decreases when user is unconscious or unable to respond
Solution Approach 1:
The system performs preliminary configuration where the user sets up fall detection preferences and emergency contacts in advance when capable. The device then automatically executes fall detection and emergency response without requiring user action during the actual fall event, ensuring both user control during setup and reliable automatic detection during emergencies
3Measurement precision
If video sensors are used for fall detection, then detection accuracy improves, but privacy concerns and system complexity increase
Solution Approach 1:
The patent extracts only the essential sensing functionality from complex video surveillance systems. Instead of using cameras that capture visual images, the system uses mobile devices with accelerometers and gyroscopes that detect motion patterns, thereby achieving fall detection accuracy while eliminating privacy concerns and reducing system complexity
4Ease of operation
If audio sensors are used for fall detection, then non-contact detection is achieved, but accuracy decreases due to environmental noise
Solution Approach 1:
The system merges multiple sensor types (accelerometer, gyroscope, proximity sensor) within the mobile device to detect falls. By combining data from these different sensors, the system achieves accurate fall detection without relying solely on audio sensors, thereby maintaining non-contact detection capability while improving accuracy through multi-sensor fusion
Data Source
AI summary
Described herein are systems and methods for automatically detecting a behavior of a monitored individual, for example, that the individual has fallen. In certain embodiments, a system is presented that includes one energy sensor (e.g., a camera of a mobile electronic device) configured to capture reflected energy (e.g., light) within a field-of view; an optional lens to modify the distance or angular range of the field-of-view; and an optional image obfuscator to blur or distort the images received by the energy sensor, thereby preserving privacy. Techniques are described for determining spatial measurements and body measurements from the images and using these measurements to identify a behavior of the monitored individual, for example, a fall.