Week 9
 

 

This week's items

1. Post results of codes
I'm very happy about this week's results. It gives us high hopes for the simulation using a wooden model. I'll see use the IQeye to see if we can detect motion of a piece of the model and capture images before and after the event. I need to look on how to automate the process of taking the picture and activate the code for recognition.
To see this week's result go to  Recognition Section

We have to start thinking on how to reduce the error due to light variations. I found a source that gives several methods to reduce the error when using eigenfaces.
Preprocessing Techniques for Eigenface Face Recognition

Since we would be looking at buildings, position is not a problem. I read that the Fisherface method is better than Eigenface. The problem with Fisherface is that it is slower. Have you though on using Fisherface as an alternative, or to add robustness to the system.

Next week is the final week of the summer term. I'll like to know what are your plans as far as my job is concerned. I'll be on vacation starting the 6th until the 21st of September.  I have to think about my options and obligations for next term. Maybe we could talk by phone if you have questions.