# Building Footprint Segmentation
#### Library to train building footprint on satellite and aerial imagery.




<a href='https://ko-fi.com/fuzailpalnak' target='_blank'><img height='36' style='border:0px;height:36px;' src='https://az743702.vo.msecnd.net/cdn/kofi1.png?v=0' border='0' alt='Buy Me a Coffee at ko-fi.com' /></a>

## Installation
pip install building-footprint-segmentation
## Dataset
- [Massachusetts Buildings Dataset](https://www.cs.toronto.edu/~vmnih/data/)
- [Inria Aerial Image Labeling Dataset](https://project.inria.fr/aerialimagelabeling/)
## Training
- [Train With Config](https://github.com/fuzailpalnak/building-footprint-segmentation/blob/main/examples/Run%20with%20config.ipynb)
, Use [config template](https://codebeautify.org/yaml-validator/cbc60637) for generating training config
- [Train With Arguments](https://github.com/fuzailpalnak/building-footprint-segmentation/blob/main/examples/Run%20with%20defined%20arguments.ipynb)
## Visualize Training
##### Test images at end of every epoch
- Follow [Example](https://github.com/fuzailpalnak/building-footprint-segmentation/blob/main/examples/TestCallBack.ipynb)
##### Visualizing on Tensorboard
```python
from building_footprint_segmentation.helpers.callbacks import CallbackList, TensorBoardCallback
where_to_log_the_callback = r"path_to_log_callback"
callbacks = CallbackList()
# Ouptut from all the callbacks caller will be stored at the path specified in log_dir
callbacks.append(TensorBoardCallback(where_to_log_the_callback))
```
To view Tensorboard dash board
tensorboard --logdir="path_to_log_callback"
## Defining Custom Callback
```python
from building_footprint_segmentation.helpers.callbacks import CallbackList, Callback
class CustomCallback(Callback):
def __init__(self, log_dir):
super().__init__(log_dir)
where_to_log_the_callback = r"path_to_log_callback"
callbacks = CallbackList()
# Ouptut from all the callbacks caller will be stored at the path specified in log_dir
callbacks.append(CustomCallback(where_to_log_the_callback))
```
## Split the images in smaller sample
```python
import glob
import os
from image_fragment.fragment import ImageFragment
# FOR .jpg, .png, .jpeg
from imageio import imread, imsave
# FOR .tiff
from tifffile import imread, imsave
ORIGINAL_DIM_OF_IMAGE = (1500, 1500, 3)
CROP_TO_DIM = (384, 384, 3)
image_fragment = ImageFragment.image_fragment_3d(
fragment_size=(384, 384, 3), org_size=ORIGINAL_DIM_OF_IMAGE
)
IMAGE_DIR = r"pth\to\input\dir"
SAVE_DIR = r"pth\to\save\dir"
for file in glob.glob(
os.path.join(IMAGE_DIR, "*")
):
image = imread(file)
for i, fragment in enumerate(image_fragment):
# GET DATA THAT BELONGS TO THE FRAGMENT
fragmented_image = fragment.get_fragment_data(image)
imsave(
os.path.join(
SAVE_DIR,
f"{i}_{os.path.basename(file)}",
),
fragmented_image,
)
```
## Segmentation for building footprint
- [x] binary
- [ ] building with boundary (multi class segmentation)
## Weight File
- [RefineNet trained on INRIA](https://github.com/fuzailpalnak/building-footprint-segmentation/releases/download/alpha/refine.zip)
- [DlinkNet trained on Massachusetts Buildings Dataset](https://github.com/fuzailpalnak/building-footprint-segmentation/releases/download/alpha/DlinkNet.zip)
## Commonly used utility task when working with Geotiff
- [Generate bitmap from shape file](https://github.com/fuzailpalnak/py-gis-utility#generate-bitmap-from-shape-file)
- [Generate shape geometry from geo reference bitmap](https://github.com/fuzailpalnak/py-gis-utility#generate-shape-geometry-from-geo-reference-bitmap)
- [Save Multi Band Imagery](https://github.com/fuzailpalnak/py-gis-utility#save-multi-band-imagery)