computer vision

Computer Vision and Modern Control UFMFRC-30-M
Assignment, 2014/15.
Submission deadline: 2 pm on 19th February 2015. On-line submission, via blackboard is required; please submit the needed files and the report as a pdf file.
Introduction
This assignment involves the development of a LabVIEW program to undertake some image processing, as well as writing a short essay relating to machine vision and image processing.
The aim of the LabVIEW part is to write a program that will take an image of a number of coins, threshold the image, and perform some particle analysis that can be used to calculate the value of the coins.
You will be provided with a sample image of some coins (called coins.bmp and coins2.bmp) Please ensure that your VI is able to work with both of these images.
The two parts of the assignment are to be undertaken and documented in a report in pdf format (maximum length 8,000 words), containing the required details (see below), as well as the following files:
1. The LabVIEW virtual instrument (VI) program you have developed. (Do not submit other people’s programs, or programs downloaded from the Internet.)
2. 2 image files, showing the results of processing the 2 images of the coins, using your VI.
Details required in the report
Include a contents, conclusion, and bibliography in the report. Include print outs of your LabVIEW VI, and include notes explaining how it works (i.e. say what the function of each sub-VI is).
The essay required in the report is to form a discussion of factors that would influence a demonstration vision system for coin identification through the processing of images of coins. Include in your discussion details of how you would recommend configuring such a system (e.g., the type of lighting, cameras required). The VI you developed only measures coin areas, however, such a system would also need to be able to differentiate between coins having very similar areas but different shapes. Discuss shape analysis algorithms that might be suitable for achieving this (in each case briefly describe how the algorithm works). Use diagrams to help with explanations. Identify any problems that you might expect to encounter and suggest
possible solutions. It is very important to ensure that your discussion relates to the coin application, rather than just being a general discussion.
LabVIEW programming
The required task involves building upon the simple VI that you developed in the LabVIEW Tutorial (Labview Image Display.vi), to achieve thresholding of the image and particle analysis for measuring the area of the coins.
Consider using the IMAQ Vision IMAQ RemoveParticleVI to remove any ‘noisy’ particles from the image using a specific number of ‘erosions’ (noise can be defined as pixel clusters that are smaller than two contiguous pixels in size, and hence using 2 successive erosion operations will remove the noisy particles from the image. Any coins that straddle the border of the image are to be neglected through use of the IMAQ RejectBorder VI. The IMAQ FILLHole VI can be used to fill any holes present in the detected blobs.
Make full use of the on-line help and the resources at www.ni.com. The details of ‘Labview Image Display.vi’ are shown in the diagrams below: