COMP9517: Computer Vision
2025 T1 Lab 4 Specification
Maximum Marks Achievable: 2.5
This lab is worth 2.5% of the total course marks.


Objective: This lab revisits important concepts covered in the Week 5 lectures and aims to
make you familiar with implementing specific algorithms.

Software: You are required to use OpenCV 3+ with Python 3+ and submit your code as a
Jupyter notebook (see coding and submission requirements below). In the tutor consultation
session this week and next, you can ask any questions you may have about this lab.

Materials: The sample images to be used in this lab are available via WebCMS3.

Submission: All code and requested results are assessable after the lab. Submit your source
code as a Jupyter notebook (.ipynb file) that includes all outputs (see coding requirements at
the end) by the above deadline. The submission link will be announced in due time.

Task (2.5 marks)
In many computer vision applications, image segmentation is a crucial first step towards
quantitative image analysis. As discussed in the lectures, there are many possible image
segmentation methods, ranging from very simple to very complicated.

Despite its simplicity and limitations, intensity thresholding is still a very popular image
segmentation method. This is because imperfections in the segmentation results can often be
filtered out by postprocessing using binary morphological operators.

The goal of this lab is to write an algorithm that performs image segmentation and subsequent
analysis using just intensity thresholding and binary morphological operators. More
specifically, the algorithm should be able to take any of the given lab images, produce a
segmentation result containing only the bigger objects that are completely contained in the
image, and print the number of objects (see the example results below).

The lab files should be submitted online.
Instructions for submission will be posted closer to the deadline.
Deadline for submission is Week 7, Friday 4 April 2025, 18:00:00.
Strawberries Leaves Pikachus




n = 21 n = 72 (light blue only) n = 20

Hints: Convert the input image to grayscale, apply intensity thresholding (using either a single
threshold or multiple to segment out intensities within a range), use binary morphology
operators to fill any holes and filter out noise pixels in the segmentation, use binary
reconstruction and subtraction to get rid of segmented objects touching the image boundary,
set the remaining objects that are smaller than some size threshold to black, and then count
and print the final number of objects.

The above example results were produced by a single algorithm. The only differences in
execution for each image were the parameter values of some steps of the algorithm.

Show the segmentation results obtained using a single algorithm for each input image in your
Jupyter notebook and include the count of objects. If your algorithm has any user parameters,
their values must be specified at the beginning of your code, so that users can easily change
and experiment with these parameters.

Coding Requirements
Make sure that in your Jupyter notebook, the input images are readable from the location
specified as an argument, and all output images and other requested results are displayed in
the notebook environment. All cells in your notebook should have been executed so that the
tutor/marker does not have to execute the notebook again to see the results.

Copyright: UNSW CSE COMP9517 Team. Reproducing, publishing, posting, distributing, or
translating this lab assignment is an infringement of copyright and will be referred to UNSW
Student Conduct and Integrity for action.
Released: 17 March 2025

学霸联盟