ColorClusteringPython

Description: This is a python code for finding color patches in an image or binary input

This is a Repository maintained and developed by Lokesh Ramina

About: The code is about color clusturing in 2D image. Taking inputs as .bin file.

Prerequisit: python2.7 should be installed python-pip should be install in the system. To install you can the command as per below instruction. # cd ../ColorClusteringPython # sudo apt install python-pip once installed # pip install -r requirements.tx

Instrution to execute file: Make sure the prerequisite is completed The main code is in ../ColorClusteringPython/main Save the .bin file in the main folder file name main.py To run the code: # cd .../ColorClusteringPython/main/dist/count-areas # ./count-areas _{--shape , >count-areas --shape , eg : count-areas data/sample.bin --shape 256,256 if everything is perfect output will be generated or else code will exit.}

---

Allgorithm explained: The logic is developed by Lokesh Ramina Considering the below matrix.

| 120  120  120  003  000|
| 120  120	120  003  003|	
| 003  120  120  120  003|
| 003  003  003  003  003|
| 003  003  003  003  003|

> create a parallel similar matrix which will have assignment of group value
> moving from 0,0 to w_,h_ it will check the position is assigned or not if not than will process the positon
> In process it checks the color value of near 8 locations and if any has same value than both gets assigned into same group and so on.
> output below of above image assign matrix

| 1  1  1  2  3|
| 1  1	1  2  2|	
| 2  1  1  1  2|
| 2  2  2  2  2|
| 2  2  2  2  2|