Interactive segmentation

User Manual

 

To initiate the program, read the readme.md and follow these steps:

 

1. Create a folder named "**Workspace_data**".

2. Place your input file named "**input.npy**" inside this folder.

3. Create an empty folder named "**Raw_Figs**".

4. Run "**main.py**".

 

**Object Creation**

 

The "**Object_creation**" file contains functions dedicated to generating an object set. Initially, the target objects are identified, cropped, and organized into an object set. This set includes various shape features extracted for each individual object. Additionally, you may need to develop your own function for detecting objects tailored to your specific dataset.

 

**Feature Space Clustering**

 

Within the "**feature_space_clustering**" file, functions are designed to cluster objects based on their shapes into different groups within the feature space. Notably, the number of clusters is automatically determined during the clustering process.

 

**Segmentation**

 

The "**Segmentation.py**" file includes all functions used for segmentation purposes. ... mehrIf desired, additional segmentation methods can be easily incorporated to evaluate their effectiveness on the objects under study.