School of Engineering

UCLan Coursework Assessment Brief

Academic Year: 2020-21

Module Title: Computer Vision

Module Code: EL3105

Level: 6

Objects Detection in Video

This assessment is worth 50% of the overall module mark

THE BRIEF/INSTRUCTIONS

This assignment is designed to give you insight into selected aspects of computer vision applied to object detection and tracking in video. You are asked to solve various tasks including the detection of keypoints and their robust matching, writing computer vision software operating in a soft real-time as well as testing your solution and interpreting the results.

This assignment will enable you to:

• Deepen your understanding of the keypoint detection and robust matching between keypoint data sets.

• Recognize software design challenges behind implementations of computer vision algorithms.

• Design and optimise software to meet specified requirements.

• Acquire a hands-on understanding of object detection in video problems.

(These correspond to point 1, 2, 4 and 5 of the module learning outcomes)

Assignment Description and Objectives

You are asked to write software to detect predefined objects in pre-recorded video sequences provided as part of this assignment. The solution should be based on the keypoints matching technique, using points detected respectively in the current video fame and the template images showing objects of interest. You should explore different design options, including the selection of the keypoint detection method, keypoint descriptor and keypoint transformation model. You should write the software using matlab.

In the first task you are asked to detect the object of interest, shown in the object_1.jpg template image, in the video_1 sequence.

For the second task, you should modify the software from the first task, so two predefined objects shown in object_1.jpg and object_2.jpg template images are simultaneously detected in each frame of video_2 sequence.

For the third task, you should further modify the software from the task two, so it can detect multiple instances of the same object in video_3 sequence. You should make sure that the software consistently tracks these multiple instances across video frames.

Images object_1.jpg and object_2.jpg showing objects to be detected/tracked and the video sequences video_1, video_2, and video_3 are all available from Blackboard.

Marking scheme

Your report should contain the following elements; it will be marked in accordance with the following marking scheme:

References:

R. Hartley, A. Zisserman, "Multiple View Geometry in Computer Vision," Cambridge University Press, 2003.

H. Bay, T. Tuytelaars, L.V. Gool, “SURF: Speed Up Robust Features”, European Conference on Computer Vision , ECCV’2006, pp. 404-417. 2006.

H. Al-Sahaf, et al., “Keypoints Detection and Feature Extraction: A Dynamic Genetic Programming Approach for Evolving Rotation-Invariant Texture Image Descriptors”, IEEE Transactions on Evolutionary Computations, Vol. 21, No. 6, pp. 825-844, 2017.

Matlab help on: “Object Detection in Cluttered Scene Using Point Feature Matching”

PREPARATION FOR THE ASSESSMENT

The assignment is to be introduced and discussed during laboratory sessions on Friday 26^th of February as well as 5^th and 12^th of March. During those session the background of this assignment will be introduced; the data structure will be explained, and the expected results will be elucidated with examples. The set of software tools available for the assignment will be also described.

All the algorithmic aspects necessary for the successful completion of the assignment were or will be covered during the lectures, tutorial, and laboratory sessions, these include keypoint detection, keypoint descriptor calculation, robust matching of the keypoints, and estimation of a transformation aligning matched keypoints.

RELEASE DATES AND HAND IN DEADLINE

Assessment Release date: 5/03/2021 Assessment Deadline Date and time: 14/04/2021 - 23:59

Please note that this is the final time you can submit – not the time to submit!

Feedback for this assessment will be provided by 05/05/2021.

SUBMISSION DETAILS

Submission of assignment work

This assignment constitutes 50% of the total module assessment mark. You should write a report for this assignment documenting your solutions for the three tasks defined above. The report should include a very brief introduction describing the problem, description of your adopted solutions, a more extensive description of the results and conclusions section summarising the results. The report should be approximately 1500 words long plus relevant materials (References and Appendices). You should use Harvard referencing system for this report. The report should be submitted electronically to Turnitin through Blackboard.

You should submit a documented matlab code solving the defined above three tasks. The code should be self-contained, i.e. it should be able to run as it is, without a need for any addition tools/libraries. You might be asked to explain operation of your software. The code should be submitted separately from the report into Blackboard EL3105 assignment area denoted as “Assignment Code and Videos”.

You are also expected to submit three videos (in avi format), one for each of the three tasks, demonstrating performance of your solution on the three videos video_1, video_2, and video_3 provided as part of this assignment. The submitted videos are the integral part of the assignment, up to 50 marks could be deduced in case these videos are missing from your submission. The videos should be submitted together with the code into Blackboard EL3105 assignment area denoted as “Assignment Code and Videos”.

Late work

Work submitted electronically may be submitted after the deadline to the same Turnitin assignment slot and will be automatically flagged as late.

Penalties for late submission

Except where an extension of the hand-in deadline date has been approved lateness penalties will be applied in accordance with University policy as follows:

(Working) Days Late Penalty

1 - 5 maximum mark that can be achieved: 40%

more than 5 0% given

Plagiarism

During the induction and via your student handbook, you were informed of the serious consequences of using or attempting to use unfair means to enhance performance. This includes plagiarism. The work submitted must be your own and any information and material used properly identified and acknowledged.

The University operates an electronic plagiarism detection service where your work may be uploaded, stored and cross-referenced against other material. The software searches the World Wide Web and extensive databases of reference material to identify duplication.

For detailed information on the procedures relating to plagiarism, please see the current version of the University Academic Regulations.

HELP AND SUPPORT

• The support for this assignment will be provided during scheduled extra session.

• For support with using library resources, please contact subject Mr. Robert Frost <[email protected]> or <[email protected]>. You will find links to lots of useful resources in the My Library tab on Blackboard.

• If you have not yet made the university aware of any disability, specific learning difficulty, long-term health or mental health condition, please complete a Disclosure Form. The Inclusive Support team will then contact to discuss reasonable adjustments and support relating to any disability. For more information, visit the Inclusive Support site.

• To access mental health and wellbeing support, please complete our online referral form. Alternatively, you can email [email protected], call 01772 893020 or visit our UCLan Wellbeing Service pages for more information.

• If you have any other query or require further support you can contact The , The Student Information and Support Centre. Speak with us for advice on accessing all the University services as well as the Library services. Whatever your query, our expert staff will be able to help and support you. For more information , how to contact us and our opening hours visit Student Information and Support Centre.

• If you have any valid mitigating circumstances that mean you cannot meet an assessment submission deadline and you wish to request an extension, you will need to apply online prior to the deadline.