| Name | image-augs JSON |
| Version |
2.4.16
JSON |
| download |
| home_page | |
| Summary | Image Augs supports Augmentation for Object Detection , Instance Segmentation and classification tasks. |
| upload_time | 2023-05-13 15:15:02 |
| maintainer | |
| docs_url | None |
| author | Souvik Saha |
| requires_python | >=3.7 |
| license | |
| keywords |
machine_learning
development
data_augmentations
|
| VCS |
|
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# Image Augmentations 🚀
***
<img src='images\logo.png'>
***
<img alt="GitHub code size in bytes" src="https://img.shields.io/github/languages/code-size/Souviksaha1998/Image_augmentations">
<img alt="GitHub code size in bytes" src="https://static.pepy.tech/personalized-badge/image-augs?period=total&units=international_system&left_color=black&right_color=brightgreen&left_text=Downloads">
<img alt="GitHub Repo stars" src="https://img.shields.io/github/stars/Souviksaha1998/image_augmentations?style=social">
<img alt="GitHub followers" src="https://img.shields.io/github/followers/Souviksaha1998?style=plastic">
***
This is a augemntation tool for Object Detection , Image classification and Instance Segmentation , it can perform 14 annotations. The important ones are rotation, affine, zooming in and out, noise, and blur. The augmentations were applied to a fraction of the data (40 - 50 percent of the images out of 100 can be augmented).When rotating or zooming in and out, the bounding box coordinates will also change as the image is rotated or zoomed.
***
## Code Starts from here
### 👩🏼💻Create a folder first, inside that folder keep your image annotation folder 👩🏼💻
***
<img src='images\3.jpg'>
### 👩🏼💻 Create a virtual environment 👩🏼💻
```python
pip install virtualenv
#name your environment
python3 -m venv <your env name>
#activate the environment --> for linux user
source <your env name>/bin/activate
#activate the environment --> for windows user
<your env name>/Source/activate.ps1
```
### Installation (for pip installation) 🚀
```python
pip install image_augs
```
## After installation 🎯
***
**Create a .py script inside your created folder**
**This Script is for OBJECT DETECTION**
```python
#import these modules in your created <scriptname>.py file
from object_detection_new.txt_reader_rect import RectAugmentation
################# image height and width combination ##################
# first combination --> for custom image size
# image_height = < custom image size > ; 640
# image_width = < custom image size > ; 320
# second combination --> keep aspect ratio of the image
# image_height = 640
# image_width = 'keep_aspect_ratio_False'
# Third combination --> keeping original image height and width
# image_height = 'keep_original_image_height'
# image_width = 'keep_original_image_width'
annotation_folder = 'your folder'
new_aug_saved_folder = 'new saved folder'
train_split = 0.90
image_H = 640 #check above for height and width setting
image_W = 'keep_aspect_ratio'
rect_aug = RectAugmentation(new_aug_saved_folder)
rect_aug.Image_augmentation(annotation_folder,
train_split=train_split,
image_height= image_H,
image_width= image_W,
blur=False, blur_f=0.8,
motionBlur= False , motionBlur_f= 0.8 ,
rotate=False, rotate_f = 0.8,
noise=False, noise_f=0.8,
perspective=False, perspective_f = 0.8,
affine=False, affine_f=0.8,
brightness=False, brightness_f=0.8,
hue=False, hue_f=0.8,
removesaturation=False, removesaturation_f=0.8,
contrast=False, contrast_f=0.8,
upflip=False, upflip_f=0.8,
shear=False, shear_f=0.8,
rotate90=False, rotate90_f =0.8,
blur_and_noise=False, blur_and_noise_f=0.8,
image_cutout = False, image_cutout_f=0.8,
mix_aug= False, mix_aug_f=0.8,
temperature_change= False, temperature_change_f=0.8, # change color temperature from cool to warm color
weather_change=True,weather_change_f=0.8), # add rain , fog , snow in your images
#results will be saved in < your given folder >
```
***
**This Script is for INSTANCE SEGMENTATION**
```python
#import these modules in your created <scriptname>.py file
from instance_seg.json_reader_poly import PolygonAugmentation
################# image height and width combination ##################
# first combination --> for custom image size
# image_height = < custom image size > ; 640
# image_width = < custom image size > ; 320
# second combination --> keep aspect ratio of the image
# image_height = 640
# image_width = 'keep_aspect_ratio_False'
# Third combination --> keeping original image height and width
# image_height = 'keep_original_image_height'
# image_width = 'keep_original_image_width'
#### yolo ####
# if yolo False then it will normalize all images and save it as txt , if false augmentations will be saved as json.
annotation_folder = 'your data'
new_aug_saved_folder = 'new saved dataset name'
train_split = 0.70
image_H = 640 #check above for height and width setting
image_W = 'keep_aspect_ratio'
yolo = True
im_aug_helper = PolygonAugmentation(aug_save_folder_name=new_aug_saved_folder,
yolo=yolo)
im_aug_helper.Image_augmentation(annotation_folder,
train_split=train_split,
image_height= image_H,
image_width= image_W,
blur=True, blur_f=0.8,
motionBlur= False , motionBlur_f= 0.5,
rotate=True, rotate_f = 0.8,
noise=True, noise_f=0.6,
perspective=True, perspective_f = 0.6,
affine=True, affine_f=0.6,
brightness=True, brightness_f=0.6,
hue=True, hue_f=0.6,
removesaturation=True, removesaturation_f=0.6,
contrast=True, contrast_f=0.6,
upflip=True, upflip_f=0.8,
shear=True , shear_f=0.7,
rotate90=True, rotate90_f =1.0,
blur_and_noise=True, blur_and_noise_f=0.6,
image_cutout = True, image_cutout_f=0.6,
mix_aug=True, mix_aug_f=0.7,
temperature_change=True, temperature_change_f=0.5,
weather_change=True,weather_change_f=0.3)
#results will be saved in < your given folder >
***
```
**This Script is for IMAGE CLASSIFICATION**
```python
#import these modules in your created <scriptname>.py file
from classification.classification_combined import ImageAugmentHelper
### PARAMS ###
source_folder = '<source folder>'
aug_saved_folder = '<augmentation saved folder>'
train_split = 0.5
image_height = 512
image_width = 512
classification_aug = ImageAugmentHelper(source_folder=source_folder,
aug_save_folder_name=aug_saved_folder,
train_split=train_split,
height=image_height,
width=image_width)
classification_aug.augmentations(
save_raw_images=True,
blur=True, blur_f=1.0,
noise=True,noise_f=1.0,
horizontalFlip=True, horizontalFlip_f=1.0,
brightness=True, brightness_f=1.0,
contrast=True, contrast_f=1.0,
hue=True, hue_f=1.0,
saturation=True, saturation_f=1.0,
zoom=True, zoom_f=1.0,
perspective=True, perspective_f=1.0,
translation=True, translation_f=1.0,
sharpen=True, sharpen_f=1.0,
randomShadow=True, randomShadow_f=1.0
)
```
***
Use github to clone [image_augmentations](https://github.com/Souviksaha1998/Image_augmentations) repo 🖥️
Use instanceSeg_aug_script.py / classification_aug_script.py / objectDetection_augScript.py according to your needs.
***
## Contributing
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
## License
[MIT](https://choosealicense.com/licenses/mit/)
Raw data
{
"_id": null,
"home_page": "",
"name": "image-augs",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "machine_learning,development,data_augmentations",
"author": "Souvik Saha",
"author_email": "ssouvik.191@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/ea/80/bf892804c081eab2c005a23e4b658d1006bd2ab562a3d67ad4c325a510a3/image_augs-2.4.16.tar.gz",
"platform": null,
"description": "\n# Image Augmentations \ud83d\ude80\n***\n<img src='images\\logo.png'>\n\n***\n<img alt=\"GitHub code size in bytes\" src=\"https://img.shields.io/github/languages/code-size/Souviksaha1998/Image_augmentations\">\n\n<img alt=\"GitHub code size in bytes\" src=\"https://static.pepy.tech/personalized-badge/image-augs?period=total&units=international_system&left_color=black&right_color=brightgreen&left_text=Downloads\">\n\n<img alt=\"GitHub Repo stars\" src=\"https://img.shields.io/github/stars/Souviksaha1998/image_augmentations?style=social\">\n<img alt=\"GitHub followers\" src=\"https://img.shields.io/github/followers/Souviksaha1998?style=plastic\">\n\n***\n\nThis is a augemntation tool for Object Detection , Image classification and Instance Segmentation , it can perform 14 annotations. The important ones are rotation, affine, zooming in and out, noise, and blur. The augmentations were applied to a fraction of the data (40 - 50 percent of the images out of 100 can be augmented).When rotating or zooming in and out, the bounding box coordinates will also change as the image is rotated or zoomed.\n\n***\n## Code Starts from here \n\n### \ud83d\udc69\ud83c\udffc\u200d\ud83d\udcbbCreate a folder first, inside that folder keep your image annotation folder \ud83d\udc69\ud83c\udffc\u200d\ud83d\udcbb\n***\n<img src='images\\3.jpg'>\n\n### \ud83d\udc69\ud83c\udffc\u200d\ud83d\udcbb Create a virtual environment \ud83d\udc69\ud83c\udffc\u200d\ud83d\udcbb\n\n```python\n\npip install virtualenv\n#name your environment\npython3 -m venv <your env name>\n\n#activate the environment --> for linux user\nsource <your env name>/bin/activate\n\n#activate the environment --> for windows user\n<your env name>/Source/activate.ps1\n\n```\n\n### Installation (for pip installation) \ud83d\ude80\n\n```python\npip install image_augs\n```\n\n## After installation \ud83c\udfaf\n***\n\n**Create a .py script inside your created folder**\n\n**This Script is for OBJECT DETECTION**\n\n```python\n#import these modules in your created <scriptname>.py file\n\n\nfrom object_detection_new.txt_reader_rect import RectAugmentation\n\n################# image height and width combination ##################\n\n# first combination --> for custom image size\n#\u00a0 image_height = < custom image size > ; 640\n#\u00a0 image_width = < custom image size > ; 320\n\n\n# second combination --> keep aspect ratio of the image\n#\u00a0 image_height = 640\n#\u00a0 image_width\u00a0 = 'keep_aspect_ratio_False'\n\n# Third combination --> keeping original image height and width\n#\u00a0 image_height = 'keep_original_image_height'\n#\u00a0 image_width = 'keep_original_image_width'\n\n\n\nannotation_folder = 'your folder'\nnew_aug_saved_folder = 'new saved folder'\ntrain_split = 0.90\nimage_H = 640 #check above for height and width setting\nimage_W = 'keep_aspect_ratio'\n\n\nrect_aug = RectAugmentation(new_aug_saved_folder)\n\nrect_aug.Image_augmentation(annotation_folder,\n \n train_split=train_split,\n image_height= image_H,\n image_width= image_W,\n\n\n blur=False, blur_f=0.8,\n\n motionBlur= False , motionBlur_f= 0.8 ,\n\n rotate=False, rotate_f = 0.8, \n\n noise=False, noise_f=0.8,\n\n perspective=False, perspective_f = 0.8,\n\n affine=False, affine_f=0.8,\n\n brightness=False, brightness_f=0.8,\n \n hue=False, hue_f=0.8,\n\n removesaturation=False, removesaturation_f=0.8,\n\n contrast=False, contrast_f=0.8,\n\n upflip=False, upflip_f=0.8,\n\n shear=False, shear_f=0.8, \n\n rotate90=False, rotate90_f =0.8,\n\n blur_and_noise=False, blur_and_noise_f=0.8,\n\n image_cutout = False, image_cutout_f=0.8,\n \n mix_aug= False, mix_aug_f=0.8, \n \n temperature_change= False, temperature_change_f=0.8, # change color temperature from cool to warm color\n\n weather_change=True,weather_change_f=0.8), # add rain , fog , snow in your images\n \n \n\n#results will be saved in < your given folder >\n```\n***\n\n**This Script is for INSTANCE SEGMENTATION**\n\n```python\n#import these modules in your created <scriptname>.py file\nfrom instance_seg.json_reader_poly import PolygonAugmentation\n\n\n################# image height and width combination ##################\n\n# first combination --> for custom image size\n#\u00a0 image_height = < custom image size > ; 640\n#\u00a0 image_width = < custom image size > ; 320\n\n\n# second combination --> keep aspect ratio of the image\n#\u00a0 image_height = 640\n#\u00a0 image_width\u00a0 = 'keep_aspect_ratio_False'\n\n# Third combination --> keeping original image height and width\n#\u00a0 image_height = 'keep_original_image_height'\n#\u00a0 image_width = 'keep_original_image_width'\n\n\n#### yolo ####\n# if yolo False then it will normalize all images and save it as txt , if false augmentations will be saved as json.\n\nannotation_folder = 'your data'\nnew_aug_saved_folder = 'new saved dataset name'\ntrain_split = 0.70\nimage_H = 640 #check above for height and width setting\nimage_W = 'keep_aspect_ratio'\nyolo = True\n\n\nim_aug_helper = PolygonAugmentation(aug_save_folder_name=new_aug_saved_folder,\n yolo=yolo)\n\nim_aug_helper.Image_augmentation(annotation_folder,\n \n train_split=train_split,\n image_height= image_H,\n image_width= image_W,\n\n\n blur=True, blur_f=0.8,\n\n motionBlur= False , motionBlur_f= 0.5,\n\n rotate=True, rotate_f = 0.8, \n\n noise=True, noise_f=0.6,\n\n perspective=True, perspective_f = 0.6,\n\n affine=True, affine_f=0.6,\n\n brightness=True, brightness_f=0.6,\n \n hue=True, hue_f=0.6,\n\n removesaturation=True, removesaturation_f=0.6,\n\n contrast=True, contrast_f=0.6,\n\n upflip=True, upflip_f=0.8,\n\n shear=True , shear_f=0.7, \n\n rotate90=True, rotate90_f =1.0,\n\n blur_and_noise=True, blur_and_noise_f=0.6,\n\n image_cutout = True, image_cutout_f=0.6,\n \n mix_aug=True, mix_aug_f=0.7,\n \n temperature_change=True, temperature_change_f=0.5,\n \n weather_change=True,weather_change_f=0.3)\n#results will be saved in < your given folder >\n***\n\n```\n\n**This Script is for IMAGE CLASSIFICATION**\n\n```python\n#import these modules in your created <scriptname>.py file\nfrom classification.classification_combined import ImageAugmentHelper\n\n\n### PARAMS ###\nsource_folder = '<source folder>'\naug_saved_folder = '<augmentation saved folder>'\ntrain_split = 0.5\nimage_height = 512\nimage_width = 512\n\nclassification_aug = ImageAugmentHelper(source_folder=source_folder,\n aug_save_folder_name=aug_saved_folder,\n train_split=train_split,\n height=image_height,\n width=image_width)\n\n\nclassification_aug.augmentations(\n\n save_raw_images=True,\n\n blur=True, blur_f=1.0,\n\n noise=True,noise_f=1.0,\n\n horizontalFlip=True, horizontalFlip_f=1.0,\n\n brightness=True, brightness_f=1.0,\n\n contrast=True, contrast_f=1.0,\n\n hue=True, hue_f=1.0,\n\n saturation=True, saturation_f=1.0,\n\n zoom=True, zoom_f=1.0,\n\n perspective=True, perspective_f=1.0,\n\n translation=True, translation_f=1.0,\n\n sharpen=True, sharpen_f=1.0,\n \n randomShadow=True, randomShadow_f=1.0\n)\n```\n***\nUse github to clone [image_augmentations](https://github.com/Souviksaha1998/Image_augmentations) repo \ud83d\udda5\ufe0f\nUse instanceSeg_aug_script.py / classification_aug_script.py / objectDetection_augScript.py according to your needs.\n\n***\n\n\n## Contributing\nPull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.\n\n## License\n[MIT](https://choosealicense.com/licenses/mit/)\n\n\n",
"bugtrack_url": null,
"license": "",
"summary": "Image Augs supports Augmentation for Object Detection , Instance Segmentation and classification tasks.",
"version": "2.4.16",
"project_urls": null,
"split_keywords": [
"machine_learning",
"development",
"data_augmentations"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "6a070ac459cb4c02e879a5247b7414d0c108b59a81303d34bd4192d6328c6e32",
"md5": "9b76dc9a17f3acf16aee2abc792e455d",
"sha256": "dd91d160098b9b8bf2b24ee84b63b9be40323e07eb0e8bccfedec77e78f12255"
},
"downloads": -1,
"filename": "image_augs-2.4.16-py3-none-any.whl",
"has_sig": false,
"md5_digest": "9b76dc9a17f3acf16aee2abc792e455d",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 30983,
"upload_time": "2023-05-13T15:14:59",
"upload_time_iso_8601": "2023-05-13T15:14:59.641488Z",
"url": "https://files.pythonhosted.org/packages/6a/07/0ac459cb4c02e879a5247b7414d0c108b59a81303d34bd4192d6328c6e32/image_augs-2.4.16-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "ea80bf892804c081eab2c005a23e4b658d1006bd2ab562a3d67ad4c325a510a3",
"md5": "4917ef712010f2a553252dc031d7bfec",
"sha256": "f014af75a431619bb0b85d4d5fdb54c0cc772e38ae614117b3e0a31796f02a16"
},
"downloads": -1,
"filename": "image_augs-2.4.16.tar.gz",
"has_sig": false,
"md5_digest": "4917ef712010f2a553252dc031d7bfec",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 28938,
"upload_time": "2023-05-13T15:15:02",
"upload_time_iso_8601": "2023-05-13T15:15:02.176310Z",
"url": "https://files.pythonhosted.org/packages/ea/80/bf892804c081eab2c005a23e4b658d1006bd2ab562a3d67ad4c325a510a3/image_augs-2.4.16.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-05-13 15:15:02",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "image-augs"
}
