Aim
The aim of this project is design and implement a system which can detect and identify objects of
various shapes, as well as providing various metrics for each detected shape.
Procedure
Complete the following procedure and write a report to explain the theory, implementation and
debug details, and results. Submit your report together with your code before the due date. Show
the demonstrator your working system to be signed off.
Design and build a simple industry parts detection and classification system for assembly lines. The
system must be able to inspect/capture the movable objects (such as, tins, balls, cubes and others)
on the table, detect the objects with a shape specified by the user (e.g. circular shape, rectangular
shape, and triangular shape), count the number of such objects and mark their locations in the
image.
Extra Features
Consider adding one or more of the following features to your program. Those doing the project in
pairs should include at least 1-2 of these in order to be eligible for full marks in the project.
• Invariance to colour, translation, rotation, and scale.
• Texture analysis to distinguish between objects of similar shape.
• Calculation of metrics for different shapes (size, colour, aspect ratio, other parameters).
• 3D shape detection, invariant to rotation in all three dimensions (hard).
Project Option 2: Face Detection in Live Video
Aim
The aim of this project is to design and implement a machine vision system which can reliably detect
faces in a video stream.
Procedure
Complete the following procedure and write a report to explain the theory, implementation and
debugging details, and results. Submit your report together with your code before the due date. You
will also need to demonstrate your project in the last week of the couse.
Design and build a simple human face detection system for live video from the USB camera. The
system must be able to detect faces moving around in the field of view and mark their locations in
the image. (hint: use colour information for segmentation and morphological operations to
determine the shape, and other processing if necessary.) You can use the paper listed below as a
startling point, or use any other technique you think is appropriate.
N. Herodotou, K. N. Plataniotis and A. N. Venetsanopoulos, “Automatic Location and Tracking of
Facial Region in Color Video Sequence”, Signal Processing: Image Communication 14 (1999) 359-388.