Skip to content

Basic Image Processing exercises. (Trying to not use openCV implementation)

Notifications You must be signed in to change notification settings

vFones/CVexercises

Repository files navigation

Digital image processing

Some exercise for an exam. (Trying to not use openCV implementation)

How to compile

Use script.sh to build cpp files

Canny

Canny algorithm for edge detection. Pass Gaussian blur kernel, sigma, lower threshold, higher threshold.

Distance transform

Used for distance mapping.

Histogram equalizator

Equalize an image histogram.

Harris

Algorithm for Corner detection in an image.

Hough circles

Detection circles in images: minimum radius, max radius and threshold for accumulator.

Hough lines

Detection lines in images: input is the threshold for accumulator.

K-means

K-means algorithm find correct cluster starting from K centroids: applied to images find correct cluster for a pixel based on RGB distance (euclidian). Auto version stops if there isn't a variation under a certain threshold. Interactive can select with mouse centroids rather then random. Iterative use N iteration before stop.

Noise remover

Apply a median blur for salt pepper noise images or mean blur for noisy images.

Edge detection with operators

Region growing

Region growing using RGB distance and HSV tint for just a color expansion. Interactive must click on source image to expand. RGB and Greyscale version tries to expand automatically.

Split (& Merge)

Recursively split images if RGB dist is less then threshold. Try to recolor blocks up to max block size.

About

Basic Image Processing exercises. (Trying to not use openCV implementation)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published