NCTU-B4G

NCTU-B4G

Sub project 1:Sensor Device Technology

Objective: Identify different users of a luggage through Machine Learning by utilizing IMU data. Furthermore, send the result to Mobile Device in the scenario of monitoring luggage status on Public transport.

Data flow diagram:

Hierarchical KNN

1.Use first-level KNN to classify luggage’s different states(Drag、Carry、Push and  Stop).

2. According to different state, use different second-level KNN model to classify different users(5 users)

Train models for classify states and classify users

Data Input : 3-axis accelerometer, 3-axis gyroscope, Euler angle

Classify states – first level

 

Hardware:

  • Eco Mini
    • CC2541
    • MPU9250

Software:

     1.   Driver for MPU9250

     2.   EcoMini_9axisDataStream