splitraster 0.3.6

Provide good support for deep learning and computer vision tasks by creating a tiled output from an input raster dataset.

Split Raster

Provide good support for deep learning and computer vision tasks by creating a tiled output from an input raster dataset.

Find more details:

• Document: split_raster
• Pypi: splitraster

Use the packages

pip install splitraster

Try Sample code

The sample image can be found in the GitHub repo.

from splitraster import io

input_image_path = "./data/raw/RGB.png"
gt_image_path = "./data/raw/GT.png"

save_path = "../data/processed/RGB"
crop_size = 256
repetition_rate = 0.5
overwrite = False

n = io.split_image(input_image_path, save_path, crop_size,
                   repetition_rate=repetition_rate, overwrite=overwrite)
print(f"{n} tiles sample of {input_image_path} are added at {save_path}")

save_path_gt = "./data/processed/GT"
n = io.split_image(gt_image_path, save_path_gt, crop_size,
                   repetition_rate=repetition_rate, overwrite=overwrite)
print(f"{n} tiles sample of {gt_image_path} are added at {save_path_gt}")

Possible results:

Successfully installed splitraster-0.1.0
❯ python test.py
Input Image File Shape (H, W, D):(1000, 1000, 3)
crop_size=256, stride=128
Padding Image File Shape (H, W, D):(1024, 1024, 3)
There are 49 files in the ./data/processed/RGB
New image name will start with 50
Generating: 100%|█████████████████████████████████████████████████████████████| 49/49 [00:00<00:00, 50.65img/s]
49 tiles sample of ./data/raw/RGB.png are added at ./data/processed/RGB
Input Image File Shape (H, W, D):(1000, 1000)
crop_size=256, stride=128
Padding Image File Shape (H, W, D):(1024, 1024)
There are 49 files in the ./data/processed/GT
New image name will start with 50
Generating: 100%|████████████████████████████████████████████████████████████| 49/49 [00:00<00:00, 139.72img/s]
49 tiles sample of ./data/raw/GT.png are added at ./data/processed/GT

You can also work with Remote Sensing (GeoTIFF) Satellite images such as Multispectral Images which have more bands or channels. All the codes will be the same, but with a small difference. Replace the io with the geo module.

This feature also needs you to install the gdal package with the following command in your python environment. This package is not in the required packages due to many users may not use this function.

conda install -c conda-forge gdal

Sample Code:

from splitraster import geo
input_image_path = "./data/raw/Input.tif"
gt_image_path = "./data/raw/GT.tif"

save_path = "../data/processed/Input"
crop_size = 256
repetition_rate = 0.5
overwrite = False

n = geo.split_image(input_image_path, save_path, crop_size,
                   repetition_rate=repetition_rate, overwrite=overwrite)
print(f"{n} tiles sample of {input_image_path} are added at {save_path}")

Random Sampling Code

The basic implementation is still the same as the above. Just replace the 'split_image' method to 'rand_crop_image'.

from splitraster import io
input_image_path = "./data/raw/RGB.png"
gt_image_path = "./data/raw/GT.png"

input_save_path = "./data/processed/Rand/RGB"
gt_save_path = "./data/processed/Rand/GT"

n = io.random_crop_image(input_image_path, input_save_path,  gt_image_path, gt_save_path, crop_size=256, crop_number=20, img_ext='.png', label_ext='.png', overwrite=True)

print(f"{n} sample paris of {input_image_path, gt_image_path} are added at {input_save_path, gt_save_path}.")

from splitraster import geo
input_tif_image_path = "./data/raw/TIF/RGB5k.tif"
gt_tif_image_path = "./data/raw/TIF/GT5k.tif"

input_save_image_path = "./data/processed/Rand/RGB_TIF"
gt_save_image_path = "./data/processed/Rand/GT_TIF"

n = geo.random_crop_image(input_tif_image_path, input_save_image_path,  gt_tif_image_path, gt_save_image_path, crop_size=500, crop_number=20, overwrite=True)

print(f"{n} sample paris of {input_tif_image_path, gt_tif_image_path} are added at {input_save_image_path, gt_save_image_path}.")

Future Update:

• Add Random Sampling feature.
• Create a GUI with Qt and generate an executable file
• Add Sample Balancing feature.