feat(docs): add readme in exa

mple folder

README.md

@@ -30,7 +30,7 @@ Now you have everything you need to run your voxel-based pipeline.
 # Examples
 
 Examples of using the voxel-based pipeline are presented in the [`examples`](https://github.com/prime-slam/sova/tree/main/examples) 
-directory with the all necessary instructions of how to run them.
+directory with the all necessary instructions of how to run them (check also README.md).
 
 # Contributing
 

examples/README.md

@@ -0,0 +1,82 @@
+## Structure
+
+This directory contains examples of running voxel-based pipeline and scripts for visualizing plane segmentation results.
+
+```shell
+examples
+├── configurations
+│   ├── hilti.yaml
+│   └── kitti.yaml
+├── pipeline.py
+└── segmentation.py
+```
+
+Every `*.py` files contains self-documented instructions of how to run them
+
+## Pipeline configuration
+
+SOVA allows you to configure your pipeline using yaml files, which has following structure:
+```yaml
+dataset:
+  type: "hilti"             # Type of dataset reader to use
+  path: "evaluation/hilti"  # Path to data
+  patches:                  # Block for patches parameters configuration
+    start: 0                # Beginning of optimisations sequence
+    end: 29                 # End of optimisations sequence
+    step: 10                # Step of optimisations patch
+    iterations: 1           # Number of re-optimizations on single patch
+pipeline:                   # Block for pipeline configuration
+  grid:                     # Grid configuration
+    ...
+  subdividers:              # Subdividers configuration
+    ...
+  segmenters:               # Segmenters configuration
+    ...
+  backend:                  # Backend configuration
+    ...
+  output: "output/hilti"    # Path to voxel-based pipeline output
+debug: false                # Debug parameter which will be send throw the whole pipeline
+                            # if it sets true, visualizations using k3d will be saved
+```
+
+Grid configuration example:
+```yaml
+grid:
+  voxel_edge_length: 8
+```
+
+Subdividers configuration example:
+```yaml
+subdividers: # You may use one of the following conditions
+  size: 1
+  count: 50
+  eigen_value: 0.5
+```
+
+Segmenters configuration example:
+```yaml
+segmenters: # You may use one of the following conditions
+  ransac:
+    threshold: 0.01
+    initial_points: 6
+    iterations: 1000
+  cape:
+    correlation: 1
+  count:
+    count: 50
+```
+
+Backend configuration example:
+```yaml
+backend:
+  type: "eigen_factor"        # Also bareg available
+  parameters:
+    iterations_number: 5000
+    robust_type: QUADRATIC    # Also HUBER available
+```
+
+## Visualization
+
+If you use `debug: true` option, pipeline will save point clouds' k3d visualization in html files. Black colour of points mean they weren't used on optimization stage.
+
+<img src="../assets/visualization.png">