This repository contains python scripts for computer vision algorithms implementations done as part of my university course work (SBE3024) and for personal projects in one place.
The project is structured into five components,jupiter notebooks contain additional information about some of the math and details of implementations:-
- Filtration of noisy images using low pass filters such as: average, Gaussian, median.
- Edge detection using variety of masks such as: Sobel, Prewitt, and canny edge detectors.
- Histograms and equalization.
- Frequency domain filters.
- Hybrid images.
- Apply Hough transform for detecting parametric shapes like circles and lines
- Apply Active Contour Model for semi-supervised shape delineation.
- SIFT feature extraction
- Apply Harris operator and Lamda- for detecting corners.
- feature matching using SSD, NCC
- Thresholding(Otsu, local, optimum)
- Segmentation (Kmeans, agglomerative, region_growing, mean_shift)
-
dimensonality reduction using PCA
-
opencv face_cascade classifier
-
train an SVM model achieving an accuracy of 87.3%.
Some of the scripts in this repo were part of team work projects:
Team member |
---|
Ehab Kamal |
Aya Sameh |
Aya Amr |
Alaa Yasser |