Python project 2
Budget: ₹600 – ₹4,000 INR
Research and implement Harris Corner Detection using Python/Numpy. Investigate the
behaviour of the algorithm. Is it rotation invariant (consider angles up to 45o
)? Is it scaleinvariant (scales from .5 to 1.5)?
Research an appropriate region descriptor for the Harris detector and use it to find
correspondence points between pairs of images. Use these correspondences to align the
image pairs and generate a composite image.
Report on the performance of your detector using some test image pairs. Suitable test
images can be found in the Resources / Projects / Project (3): Harris Corner Detection and
Image Alignment folder on this site.
Provide your Python/Numpy code, all input images used, and a comprehensive report in
pdf. The report should cover in detail the steps you undertook to carry out the project, what
you considered to be the challenges of the project and how you overcame them, along with
a presentation of results and a discussion of same, showing the behaviour of your code,
with the aligned images shown clearly. Be sure to discuss the behaviour of the code when
working with scaled and/or rotated images. The report can be as long as you like (suggested
8-10 pages, but no upper limit).
Please note that I want both a report and your Python code files, i.e., it is not enough simply to
include the code in the report, I want it in a form that I can execute
behaviour of the algorithm. Is it rotation invariant (consider angles up to 45o
)? Is it scaleinvariant (scales from .5 to 1.5)?
Research an appropriate region descriptor for the Harris detector and use it to find
correspondence points between pairs of images. Use these correspondences to align the
image pairs and generate a composite image.
Report on the performance of your detector using some test image pairs. Suitable test
images can be found in the Resources / Projects / Project (3): Harris Corner Detection and
Image Alignment folder on this site.
Provide your Python/Numpy code, all input images used, and a comprehensive report in
pdf. The report should cover in detail the steps you undertook to carry out the project, what
you considered to be the challenges of the project and how you overcame them, along with
a presentation of results and a discussion of same, showing the behaviour of your code,
with the aligned images shown clearly. Be sure to discuss the behaviour of the code when
working with scaled and/or rotated images. The report can be as long as you like (suggested
8-10 pages, but no upper limit).
Please note that I want both a report and your Python code files, i.e., it is not enough simply to
include the code in the report, I want it in a form that I can execute