Initial Team Set Up

Project Selection

Lecturer, Jane Courtney, made accessible to us a list of projects for us to choose from. Straight away, Vehicle Recognition caught my eye as a project I would be interested in talking. This is due to the fact that I am currently proposing to undertake a final year project that deals with image recognition.

Team Meeting

Our first port of call was to actually seek each other out, and put a face to the name for each other. We began to establish the problem task at hand, try to break down what is being asked of us:

• Recognise a Vehicle in the image
• Identify key features to extract from the vehicle
• Establish a methodology of implementing the recognition autonomously

Following on from the meeting, the above mentioned items became our action items for all team members collectively.

Implementation

Next we spoke ad-lib about how we should move forward trying to implement this project. Diarmuid argued the point of using an "if-else" decision tree to match unique features of vehicle in order to determine what category vehicle a given image belonged to. Secondly, he argued the point of superimposing images of the same category of vehicle in order to find matching features.

I offered a point of using guided machine learning, with the help of some pre processing in order to categorise the vehicle, however, as pointed out by the rest of the team, this method is too much a part of a machine learning module, rather than an image processing module.

Minutes

Please see team minutes found here hosted on Google Drive (DIT sign in required) Team Meetings