This repository is the storehouse of all the codes that we have written for the course assignments of Advanced Digital Image Processing and Computer Vision (CS60052) at IIT Kharagpur, for the session of Spring 2020.
Here a brief outline of the main algorithms implemented is given.
- Scaled Bilateral Filtering. [Reference Paper]
- Harris Corner Detection.
- Morphological Operations.
- Vanishing line of an image (or a view) in the 2D homogenous space P2 by finding intersections of pairs of parallel lines.
- Affine Rectification of an image by finding the homography that sends the vanishing line to the line at infinity.
- Matching similar feature points (using SIFT) in two images of the same scene captured from separate views.
- Finding the fundamental matrix (F), camera parameters and epipoles & epipolar lines.
- Estimating (relative) 3D depth of the matched feature points in the scene.
- Finding the dominant color in a RGB color image using K-Means Clustering in the x-y 2D chromaticity plane.
- Transferring the dominant color to another image via color space transformations. [Reference Paper]
- Finding curvatures (mean, gaussian and principal) and associated topologies at each point of range data.
- Computation of Neighborhood Plane Set (NPS) at each pixel.
- Region Growing of homogenous labels using principal and gaussian curvatures and NPS. [Reference Papers]
A detailed description of the functions and methods implemented along with running instructions and requirements info can be found in the respective documentation files (pdf) inside each folder. As an example, this is the documentation of the third assignment (Stereogeometry).
[Equal contributions for assignments 1 to 3. The last two had individual submissions.]