fall detection slideshow

6
Fall Detection Student ID: 04742482 18/05/2015

Upload: anilramnanan

Post on 22-Jul-2015

18 views

Category:

Software


0 download

TRANSCRIPT

Page 1: Fall detection slideshow

Fall DetectionStudent ID: 04742482

18/05/2015

Page 2: Fall detection slideshow

What is the problem ?

• People who fall need to contact someone

• More people have smartphones

• Use technology to detect fall and alert someone

Page 3: Fall detection slideshow

Propose Solution• Application gathers accelerometer data and detects if the phone

has fallen

• Audibly alerts the user and asks if they have actually fallen ?

• Possible Scenarios

• Person answers “yes”, the app then alerts contact

• Person answers “no”, the app cancels alert

• Person does not answer or interact with phone. App assumes person in incapacitated and sends alert.

• Person can also interact via “Yes” and “No” buttons.

Page 4: Fall detection slideshow

Technical Components• Detecting a Fall

• Accelerometer data from X, Y and Z Axis. Vector is calculated using

• Vector of 9.8 m/s is at rest. Falling is 3.0 m/s

• Voice Interaction

• Android Text to Speech is used for audible alerts. android.speech.tts.TextToSpeech

• Android Speech Recognizer used to implement voice commands. Recognises “Yes” and “No” for actions. android.speech.RecognizerIntent

Page 5: Fall detection slideshow

DEMO

Page 6: Fall detection slideshow

Future Work

• Reduce False Positive Use additional sensors (Gravity, Step Counter)

• More complete voice interaction

• Reduce battery consumption

• Implement on wearable devices