October 22, 2005

Calibration of the first pair of Sharp IR sensors

Long and WB obtained a pair of the Sharp sensors from Stuart for calibration because this is the only type of sensor on the robot that has not been calibrated..or something to this effect according to Rod.

A problem faced by the technicians was that the output of this sensor gives ambigious readings. In other words, the same output voltage can be a result of 2 possible distance readings. We believe this problem has been resolved by careful reading of the datasheet. The output voltage of the sensor peaks at roughly 20cm. However the optimum range for this sensor is 20cm to 150cm which basically means that we can throw away the part of the output graph for before the peak value. This hypothesis has been confirmed by a member of the Sensors Research Group over at Engineering.

Another problem faced in the calibration was that the output voltage fluctuated greatly. This was found to be the result of the input voltage not being in the optimum range. A simple change of AA batteries temporaily resolved this issue. However, we would need to work on getting a stable voltage into the sensors, perhaps by the use of a power supply regulator as suggested by the Sensors Research Group.

The vertical sensor does not give a stable output during calibration although the horizontal one does. this will be raised to Rod and Stuart during our next meeting.

Pictures of the calibration setup can be found at:

September 15, 2005

Incomplete Break–down of this project

This is by no means final!

Please comment on it, completly change it if you want, suggest additions etc

Robot Vision
CIMG Library for image processing

  • Recognise Objects?

  • Map Suroundings to form a map of the world

  • Recognise Faces

    • Clean up image

      • remove noise

      • equalise histogram

    • Low Level Analysis

      • Edge detection

      • Thinning algorthim

      • Find oval/face shape object

    • High level feature dectection

      • eyes

      • nose

      • mouth

    • Follow people (gait analysis)

      • look at body shape

      • follow changes in movement

Robot Movement

  • Set a standard api to move the robot

    • functions like

      • Follow(object x)

      • Stop()

      • Move(for time x)

      • Move(for distance x)

      • Turn(degree x)

      • etc.....

well sorted out the multi level bullet point problem but now i have these massive gaps between them, d'oh!

The MoboRobo Blog

Hi welcome to the MoboRobo blog!

October 2022

Mo Tu We Th Fr Sa Su
Sep |  Today  |
               1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

Search this blog


Most recent comments

  • Hey, we need to meet up to discuss the schedule by on this entry
  • One for the team by on this entry

Blog archive

RSS2.0 Atom
Not signed in
Sign in

Powered by BlogBuilder