Python Version 3.7
This project stitches images together to create a Panorama. Images are taken from different perspectives and transformed to fit together by using SIFT feature detection and RANSAC algorithm for homography transformations.
-
Ensure that your Python version is 3.7 and pip is installed
-
Open the project folder (Project_RANSAC) as a Python project in your choice of IDE (i.e. PyCharm)
Set the working directory is set to/Project_RANSAC/src
-
Select Python 3.7 as your interpreter
OpenCV Contrib v.3.4.2.16 files are found in the project folder.- Required Packages
- numpy
pip install numpy
- OpenCV Contrib v.3.4.2.16
- 64-bit: opencv_contrib_python-3.4.2.16-cp37-cp37m-win_amd64.whl
pip install opencv_contrib_python-3.4.2.16-cp37-cp37m-win_amd64.whl
- 32-bit: opencv_contrib_python-3.4.2.16-cp37-cp37m-win32.whl
pip install opencv_contrib_python-3.4.2.16-cp37-cp37m-win32.whl
- 64-bit: opencv_contrib_python-3.4.2.16-cp37-cp37m-win_amd64.whl
- numpy
- Required Packages
- Harris Corner Response threshold, RANSAC Iterations and RANSAC threshold can be specified in the global variables found in
Main.py
(Line 6-8)
- Run
Main.py
to execute the program - To go to the next image, press any key (excluding Esc). Output images are saved in
src/project_images/output
The sequence of the images are as follows:- Boxes Harris Corner Detection
Rainier1 Harris Corner Detection
Rainier2 Harris Corner Detection - SIFT matches for Rainier 1-2
- RANSAC Matches for Rainier1 and Rainier2
- Stitched Rainier1 and Rainier2 (R1-2)
- RANSAC Matches for R1-2 and Rainier3 (R1-3)
- RANSAC Matches for R1-3 and Rainier4 (R1-4)
- RANSAC Matches for R1-4 and Rainier5 (R1-5)
- RANSAC Matches for R1-5 and Rainier6 (R1-6)
- Stitched Rainier1-6 (Scaled 50% For Displaying)
- RANSAC Matches for Japan1 and Japan2 (JP1-2)
- RANSAC Matches for JP1-2 and Japan3 (JP1-3)
- Stitched Japan1-3
- RANSAC Matches for Hanging1 and Hanging1-2
- Stitched Hanging1-2
- Boxes Harris Corner Detection