This page records the current status of e-log archives with technical analysis.

Data collection

ViconRevue

ViconRevue is the wearable digital camera designed to take photos passively without intervention whilst being worn by the user. Based on Microsoft SenseCam technology, Revue is a research tool aimed at researchers as an aid for people with memory loss.

Specification

Vicon Revue structure -- Screenshot from the ViconRevue manual

Vicon Revue contains a 1230mAh lithium-ion battery that lasts enough for one-day continuous shot. It is built around a PIC microcontroller. It has:

  • 1GB of flash memory
  • 4KB RAM
  • Many general purpose I/O (GPIO) lines
  • On-chip peripherals.

Vicon Revue stores VGA resolution (640x480) images as compressed .JPG files on the 1GB flash memory card. The typical image size (around 30KB) allows over 30,000 photos to be stored in the camera. ViconRevue has the wide-angle lens with indicator lights and buttons for manual shot and privacy pause. It embeds multiple sensors that tell much details on the environment surrounding the user:

  • ACC = 3 axis accelerometer data, the accelerometer triggers the camera but doesn’t get used by the software at this time.
  • TMP = Temperature in degrees C.
  • CLR = Light level, big number (3,000+) = lots of light.
  • BAT = Battery level in millivolts, max is around 41000.
  • MAG = Magnetometer readings in 3 axis. This can be turned into a 'compass' reading but I don’t have any details on this at the moment.
  • PIR = Passive infrared sensor. 0 or 1 where 1 means that there is some moving IR source (typically a person) in view.

Three letter acronyms above are the ID of the sensing records stored in the ViconRevue memory. Records consist of image files and SENSOR.CSV text file. SENSOR.CSV include the timestamped sensing data of image and above sensors.

To get some sample videos composed of pictures captured with Revue, visit ViconRevue YouTube channel[1]. One sample video is copied here.

iPhone 3GS

I use iPhone to capture the location information using MotionX GPS app. See the example below.

Trento

GMail

Programmed using PHP for GMail IMAP acccess to monitor all email activities.

Phone records

Collecting from the phone bill.

People identification

Currently using PICASA to identify people face and classification from ViconRevue images. I am considering using Face.com open API for automated face recognition.

Status

As of October 29th, 2010 since July 9th, 2010.

  • ViconRevue Sensing records: 1,820,779
  • Images: 327,542
  • GPS points: 32,473
  • Human faces: 7,658
  • Idenfitied people: 126
  • E-mails: 8,337
  • Phone records: 243
  1. http://www.youtube.com/user/ViconRevue