Components

Intel Realsense D435i camera

*Active IR technology

*High resolution RGB camera

*Gyroscope include

*SDK for most programming language

*USB 3.1 which requires a high-level micro controller like Raspberry Pi

Nema 17 Step Motor

  • 1.5A to 2A current per phase


  • 1-4 volts


  • 3 to 8 mH inductance per phase


  • 44 N·cm (62oz·in, 4.5kg·cm) or more holding torque


  • 1.8 or 0.9 degrees per step (200/400 steps/rev respectively)


There we can choose Jetson or Raspberry Pi in our system, in the introduction, green fonts represent strengths, red fonts represent weaknesses

Jetson

  • •GPU computation

  • •Fast object detection

  • •Our own edge detection method can be implemented

  • •No need to skip any point. This makes our program more robust.

  • •Need tape covering the whole arm

  • Need two colors of tape

Raspberry Pi

•Direct I/O

•Detect arm with small area of green tape

•CPU intensive

•Slow Performance

•Need to skip points for performance


STM 32, Driver, and Motor