/COMP9517-computer-vision-k-means-thresholding

Primary LanguagePython

k-means-thresholding

Python 3+

The following libraries are also available: OpenCV 3.0+ Numpy, Numba

file structure:

  1. k_means_thresholding.{py,cpp}
  2. filtered_k_means_thresholding. {py,cpp}
  3. step_2.txt
  4. count_eggs_4. {py,cpp}

step1: thresholding of the greyscale image using K-means algorithm

● Input: a greyscale image

● Output: a binary image

● Testing sequence: python3 k_means_thresholding.py input_image_1.jpg output_binary.jpg

step2: filtering and thresholding of the greyscale image

● Input: a greyscale image

● Output: a binary image

● Testing sequence: python3 filtered_k_means_thresholding.py input_image_1.jpg output_binary.jpg

step3: partitioning similar neighbourhood pixels in output of Step 2 using connected components

● Inputs: a binary image, a positive integer n which is the minimum number of pixels in the component

● Outputs: the number of eggs with area larger than n pixels in the image written to stdout AND write a colour image to ./output_eggs.jpg where each egg found is a different colour.

● Testing sequence: python3 count_eggs_4.py 50 output_binary.jpg labeled.jpg

input image:

alt text

grey image (after step 1):

alt text

labeled image:

alt text