A collection of information about Data Augmentation libraries for Image Processing.
last updated on 10/10/2023
Transformations | Description | Tags | |||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1. | from Adds a value to all pixels in an image. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
2. | from Adds values to the pixels of images with possibly different values for neighbouring pixels. |
||||||||||||||||||||||||||
3. | from Adds to the brightness channels of input images. This is a wrapper around |
||||||||||||||||||||||||||
4. | from Add random values to the hue of images. The augmenter first transforms images to HSV colorspace, then adds random values to the H channel and afterwards converts back to RGB. If you want to change both the hue and the saturation, it is recommended to use |
||||||||||||||||||||||||||
5. | from Increases or decreases hue and saturation by random values. The augmenter first transforms images to HSV colorspace, then adds random values to the H and S channels and afterwards converts back to RGB. |
||||||||||||||||||||||||||
6. | from Add random values to the saturation of images. The augmenter first transforms images to HSV colorspace, then adds random values to the S channel and afterwards converts back to RGB. If you want to change both the hue and the saturation, it is recommended to use |
||||||||||||||||||||||||||
7. | from Adds noise sampled from gaussian distributions elementwise to images. |
||||||||||||||||||||||||||
8. | from Adds noise sampled from laplace distributions elementwise to images. |
||||||||||||||||||||||||||
9. | from Adds noise sampled from poisson distributions elementwise to images. |
||||||||||||||||||||||||||
10. | from Applies affine transformations to images. |
||||||||||||||||||||||||||
11. | from Applies PIL-like affine transformations to images. |
||||||||||||||||||||||||||
12. | from Applies CLAHE to all channels of images in their original colorspaces. CLAHE (Contrast Limited Adaptive Histogram Equalization) performs histogram equilization within image patches, i.e. over local neighbourhoods. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
13. | from Applies histogram equalization to all channels of images in their original colorspaces. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
14. | from Adjusts contrast by cutting off p% of lowest/highest histogram values. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
15. | from Blurs an image by computing simple means over neighbourhoods. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
16. | from |
||||||||||||||||||||||||||
17. | from Blurs/Denoises an image using a bilateral filter. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
18. | from Alpha-blends two image sources using an alpha/opacity value. |
||||||||||||||||||||||||||
19. | from Blends images from two branches based on areas enclosed in bounding boxes. |
||||||||||||||||||||||||||
20. | from Blends images from two branches according to a checkerboard pattern. |
||||||||||||||||||||||||||
21. | from Alpha-blends two image sources using alpha/opacity values sampled per pixel. |
||||||||||||||||||||||||||
22. | from Alpha-blends two image sources using frequency noise masks. |
||||||||||||||||||||||||||
23. | from Blends images from two branches along a horizontal linear gradient. |
||||||||||||||||||||||||||
24. | from Alpha-blends two image sources using non-binary masks generated per image. |
||||||||||||||||||||||||||
25. | from Blends images from two branches according to a regular grid. |
||||||||||||||||||||||||||
26. | from Blends images from two branches based on segmentation map ids. |
||||||||||||||||||||||||||
27. | from Alpha-blends two image sources using simplex noise alpha masks. |
||||||||||||||||||||||||||
28. | from Blends images from two branches using colorwise masks. |
||||||||||||||||||||||||||
29. | from Blends images from two branches along a vertical linear gradient. |
||||||||||||||||||||||||||
30. | from Wrapper around |
PHOTOMETRY | |||||||||||||||||||||||||
31. | from Applies CLAHE (Contrast Limited Adaptive Histogram Equalization) to L/V/L channels in HLS/HSV/Lab colorspaces. CLAHE is a form of histogram equalization that normalizes within local image patches. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
32. | from Converts the style of images to a more cartoonish one. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
33. | from Crops images equally on all sides until they reach an aspect ratio. |
||||||||||||||||||||||||||
34. | from Takes a crop from the center of each image. |
||||||||||||||||||||||||||
35. | from Crops images equally on all sides until H/W are multiples of given values. |
||||||||||||||||||||||||||
36. | from Crops images equally on all sides until H/W is a power of a base. |
||||||||||||||||||||||||||
37. | from Crops images equally on all sides until their height/width are identical. |
||||||||||||||||||||||||||
38. | from Pads images equally on all sides until H/W matches an aspect ratio. |
||||||||||||||||||||||||||
39. | from Pads images equally on all sides up to given minimum heights/widths. |
||||||||||||||||||||||||||
40. | from Pads images equally on all sides until H/W are multiples of given values. |
||||||||||||||||||||||||||
41. | from Pads images equally on all sides until H/W is a power of a base. |
||||||||||||||||||||||||||
42. | from Pads images equally on all sides until their height & width are identical. |
||||||||||||||||||||||||||
43. | from Changes the temperature to a provided Kelvin value. Low Kelvin values around 1000 to 4000 will result in red, yellow or orange images. Kelvin values around 10000 to 40000 will result in progressively darker blue tones. |
||||||||||||||||||||||||||
44. | from Randomizes the order of channels in input images. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
45. | from Adds a single layer of clouds to an image. |
WEATHER | |||||||||||||||||||||||||
46. | from Adds clouds to images. This is a wrapper around |
WEATHER | |||||||||||||||||||||||||
47. | from Sets rectangular areas within images to zero. |
||||||||||||||||||||||||||
48. | from Replaces rectangular areas in images with black-ish pixel noise. |
||||||||||||||||||||||||||
49. | from Replaces rectangular areas in images with white-ish pixel noise. |
||||||||||||||||||||||||||
50. | from Replaces rectangular areas in images with white/black-ish pixel noise. |
||||||||||||||||||||||||||
51. | from Wrapper around |
PHOTOMETRY | |||||||||||||||||||||||||
52. | from Crops images. |
||||||||||||||||||||||||||
53. | from Crops/pads images by pixel amounts or fractions of image sizes. |
||||||||||||||||||||||||||
54. | from Crops images until their width/height matches an aspect ratio. |
||||||||||||||||||||||||||
55. | from Crops images down to a fixed maximum width/height. |
||||||||||||||||||||||||||
56. | from Crops images down until their height/width is a multiple of a value. |
||||||||||||||||||||||||||
57. | from Crops images until their height/width is a power of a base. |
||||||||||||||||||||||||||
58. | from Crops images until their width and height are identical. |
||||||||||||||||||||||||||
59. | from Fills one or more rectangular areas in an image using a fill mode. See paper “Improved Regularization of Convolutional Neural Networks with Cutout” by DeVries e Taylor. In contrast to the paper, this implementation also supports replacing image sub-areas with gaussian noise, random intensities or random RGB colors. It also supports non-squared areas. |
||||||||||||||||||||||||||
60. | from Wrapper around |
||||||||||||||||||||||||||
61. | from Sets a certain fraction of pixels in images to zero. |
||||||||||||||||||||||||||
62. | from Drops random channels from images. |
||||||||||||||||||||||||||
63. | from Wrapper around |
||||||||||||||||||||||||||
64. | from Transforms images by moving pixels locally around using displacement fields. See Simard, Steinkraus and Platt, Best Practices for Convolutional Neural Networks applied to Visual Document Analysis in Proc. of the International Conference on Document Analysis and Recognition, 2003. |
||||||||||||||||||||||||||
65. | from Embosses images and overlays the result with the original image. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
66. | from Changes the brightness of images. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
67. | from Converts images to grayscale. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
68. | from Changes the contrast of images. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
69. | from Changes the sharpness of images. |
SHARPNESS | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
70. | from Equalizes the image histogram. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
71. | from Converts non-snowy landscapes to snowy ones. It expects to get an image that roughly shows a landscape. |
WEATHER | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
72. | from Applies a blur filter kernel to images. |
BLUR | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
73. | from Applies a detail enhancement filter kernel to images. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
74. | from Applies an edge enhance filter kernel to images. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
75. | from Applies a strong edge enhancement filter kernel to images. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
76. | from Applies a sharpening filter kernel to images. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
77. | from Applies a smoothening filter kernel to images. |
BLUR | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
78. | from Applies a strong smoothening filter kernel to images. |
BLUR | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
79. | from Flips/mirrors input images horizontally. |
BASIC GEOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
80. | from Flips/mirrors input images vertically. |
BASIC GEOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
81. | from Wrapper around |
||||||||||||||||||||||||||
82. | from Adds fog to images. This is a wrapper around |
WEATHER | |||||||||||||||||||||||||
83. | from Wrapper around |
||||||||||||||||||||||||||
84. | from Adjusts image contrast by scaling pixel values to |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
85. | from Blurs images using gaussian kernels. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
86. | from Wrapper around |
||||||||||||||||||||||||||
87. | from Wrapper around |
||||||||||||||||||||||||||
88. | from Wrapper around |
||||||||||||||||||||||||||
89. | from Converts images to their grayscale versions. |
||||||||||||||||||||||||||
90. | from Applies histogram equalization to L/V/L channels of images in HLS/HSV/Lab colorspaces. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
91. | from Adds impulse noise to images. |
||||||||||||||||||||||||||
92. | from Wrapper around |
||||||||||||||||||||||||||
93. | from Inverts all values in images, i.e. sets a pixel from value |
||||||||||||||||||||||||||
94. | from Moves cells within images similar to jigsaw patterns. |
||||||||||||||||||||||||||
95. | from Degrades the quality of images by JPEG-compressing them. |
||||||||||||||||||||||||||
96. | from Wrapper around |
WORSENING | |||||||||||||||||||||||||
97. | from Quantizes colors using k-Means clustering. This “collects” the colors from the input image, groups them into k clusters using k-Means clustering and replaces the colors in the input image using the cluster centroids. |
||||||||||||||||||||||||||
98. | from Resizes images back to their input sizes after applying child augmenters. |
||||||||||||||||||||||||||
99. | from Adjust contrast by scaling each pixel to |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
100. | from Adjusts image contrast by scaling pixels to |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
101. | from |
||||||||||||||||||||||||||
102. | from Applies a pyramidic mean shift filter to each image. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
103. | from Blurs an image by computing median values over neighbourhoods. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
104. | from |
||||||||||||||||||||||||||
105. | from |
||||||||||||||||||||||||||
106. | from Blurs images in a way that fakes camera or object movements. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
107. | from Wrapper around |
||||||||||||||||||||||||||
108. | from Multiplies all pixels in an image with a specific value, thereby making the image darker or brighter. |
||||||||||||||||||||||||||
109. | from Multiplies and adds to the brightness channels of input images. This is a wrapper around |
||||||||||||||||||||||||||
110. | from Multiplies the brightness channels of input images. This is a wrapper around |
||||||||||||||||||||||||||
111. | from Multiplies values of pixels with possibly different values for neighbouring pixels, making each pixel darker or brighter. |
||||||||||||||||||||||||||
112. | from Multiplies the hue of images by random values. The augmenter first transforms images to HSV colorspace, then multiplies the pixel values in the H channel and afterwards converts back to RGB. This augmenter is a shortcut for |
||||||||||||||||||||||||||
113. | from Multipies hue and saturation by random values. The augmenter first transforms images to HSV colorspace, then multiplies the pixel values in the H and S channels and afterwards converts back to RGB. This augmenter is a wrapper around |
||||||||||||||||||||||||||
114. | from Multiplies the saturation of images by random values. The augmenter first transforms images to HSV colorspace, then multiplies the pixel values in the H channel and afterwards converts back to RGB. This augmenter is a shortcut for |
||||||||||||||||||||||||||
115. | from Pads images. |
||||||||||||||||||||||||||
116. | from Pads images until their width/height matches an aspect ratio. |
||||||||||||||||||||||||||
117. | from Pads images to minimum width/height. |
||||||||||||||||||||||||||
118. | from Pads images until their height/width is a multiple of a value. |
||||||||||||||||||||||||||
119. | from Pads images until their height/width is a power of a base. |
||||||||||||||||||||||||||
120. | from Pads images until their height and width are identical. |
||||||||||||||||||||||||||
121. | from Replaces pixels in images with pepper noise, i.e. black-ish pixels. |
||||||||||||||||||||||||||
122. | from Applies random four point perspective transformations to images. |
BASIC GEOMETRY | |||||||||||||||||||||||||
123. | from Applies affine transformations that differ between local neighbourhoods. |
||||||||||||||||||||||||||
124. | from Wrapper around |
||||||||||||||||||||||||||
125. | from Quantizes each array component to |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
126. | from Adds rain to images. |
WEATHER | |||||||||||||||||||||||||
127. | from Adds a single layer of falling raindrops to images. |
WEATHER | |||||||||||||||||||||||||
128. | from Applies RandAugment to inputs as described in the corresponding paper. See Cubuk et al. RandAugment: Practical automated data augmentation with a reduced search space. |
||||||||||||||||||||||||||
129. | from Samples Voronoi cells from regular grids and color-averages them. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
130. | from Samples Voronoi cells from image-dependent grids and color-averages them. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
131. | from Decreases the saturation of images by varying degrees. This creates images looking similar to |
||||||||||||||||||||||||||
132. | from Replaces pixels in an image with new values. |
||||||||||||||||||||||||||
133. | from Resizes images to specified heights and widths. |
||||||||||||||||||||||||||
134. | from Rotates images clockwise by multiples of 90 degrees. This could also be achieved using |
||||||||||||||||||||||||||
135. | from Applies affine rotation on the y-axis to input data. This is a wrapper around |
BASIC GEOMETRY | |||||||||||||||||||||||||
136. | from Replaces pixels in images with salt noise, i.e. white-ish pixels. |
||||||||||||||||||||||||||
137. | from Replaces pixels in images with salt/pepper noise (white/black-ish colors). |
||||||||||||||||||||||||||
138. | from Wrapper around |
PHOTOMETRY | |||||||||||||||||||||||||
139. | from Applies affine scaling on the x-axis to input data. This is a wrapper around |
||||||||||||||||||||||||||
140. | from Applies affine scaling on the y-axis to input data. This is a wrapper around |
||||||||||||||||||||||||||
141. | from Sharpens images and overlays the result with the original image. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
142. | from Applies affine shear on the x-axis to input data. This is a wrapper around |
BASIC GEOMETRY | |||||||||||||||||||||||||
143. | from Applies affine shear on the y-axis to input data. This is a wrapper around |
BASIC GEOMETRY | |||||||||||||||||||||||||
144. | from Wrapper around |
||||||||||||||||||||||||||
145. | from Adjust image contrast to |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
146. | from Wrapper around |
||||||||||||||||||||||||||
147. | from Adds falling snowflakes to images. This is a wrapper around |
WEATHER | |||||||||||||||||||||||||
148. | from Adds a single layer of falling snowflakes to images. |
WEATHER | |||||||||||||||||||||||||
149. | from Inverts all values above a threshold in images. |
||||||||||||||||||||||||||
150. | from Inverts all pixel values above a threshold. |
PHOTOMETRY | |||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
151. | from Wrapper around |
||||||||||||||||||||||||||
152. | from Wrapper around |
||||||||||||||||||||||||||
153. | from Completely or partially transforms images to their superpixel representation. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
154. | from Applies affine translation on the x-axis to input data. This is a wrapper around |
||||||||||||||||||||||||||
155. | from Applies affine translation on the y-axis to input data. This is a wrapper around |
||||||||||||||||||||||||||
156. | from Quantizes colors into N bins with regular distance. |
||||||||||||||||||||||||||
157. | from Quantizes images by setting 8-B bits of each component to zero. |
||||||||||||||||||||||||||
158. | from Uniformly samples Voronoi cells on images and averages colors within them. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
159. | from Averages colors of an image within Voronoi cells. |
||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
160. | from Applies child augmenters to brightness-related image channels. |
||||||||||||||||||||||||||
161. | from Applies child augmenters within a specific colorspace. |
||||||||||||||||||||||||||
162. | from Applies child augmenters to hue and saturation channels. |
||||||||||||||||||||||||||
163. | from Applies other augmenters in a polar-transformed space. |
||||||||||||||||||||||||||
164. | from Wrapper around |