Sourcify
Effortlessly find and manage open-source dependencies for your projects.

A powerful, stochastic image augmentation library for deep learning and computer vision.

imgaug is a highly flexible Python library designed to assist machine learning engineers in augmenting image data for deep learning architectures. Unlike basic image processing libraries, imgaug allows for the simultaneous augmentation of images, bounding boxes, keypoints, and segmentation maps, ensuring that labels remain perfectly synchronized with visual transformations. Its technical architecture is built around the concept of 'Augmenters'—stochastic objects that define probability distributions for various transformations like Gaussian noise, affine transforms, and atmospheric effects. In the 2026 market, while GPU-accelerated libraries like Albumentations have gained traction for speed, imgaug remains a preferred choice for complex research environments and legacy pipelines due to its robust support for rare data formats and highly readable 'Sequential' pipeline API. It excels in scenarios where multi-modal data (e.g., thermal + RGB) must undergo identical random transformations. The library is highly extensible, allowing developers to define custom augmenters that integrate with existing NumPy and OpenCV workflows, making it an essential component for training high-accuracy models in autonomous driving, medical imaging, and satellite analysis.
imgaug is a highly flexible Python library designed to assist machine learning engineers in augmenting image data for deep learning architectures.
Explore all tools that specialize in stochastic augmentation. This domain focus ensures imgaug delivers optimized results for this specific requirement.
Simultaneously transforms images and their corresponding bounding boxes, heatmaps, and keypoints with mathematical precision.
Allows defining parameters as probability distributions (e.g., Normal, Poisson) rather than fixed values.
A wrapper that applies a list of augmenters in a specific order or only to a percentage of the images.
Includes specialized augmenters for Fog, Clouds, Snow, and Rain using advanced noise algorithms.
Support for augmentations in various color spaces including RGB, HSV, and Lab.
Enables complex blending of original and augmented images using alpha masks.
Native support for multiprocessing to speed up augmentation on large-scale datasets.
Install via pip using 'pip install imgaug' in a Python 3.8+ environment.
Import the 'imgaug.augmenters' module, conventionally aliased as 'iaa'.
Load your source image dataset as a NumPy array of shape (N, H, W, C).
Define an Augmenter sequence using 'iaa.Sequential' to chain multiple transformations.
Configure stochastic parameters (e.g., iaa.Affine(rotate=(-25, 25))) to define ranges rather than fixed values.
Instantiate 'KeypointsOnImage' or 'BoundingBoxesOnImage' objects to track labels.
Use 'aug_det = seq.to_deterministic()' if you need to apply the same random parameters to different data types.
Execute the augmentation using 'seq.augment_images()' or 'seq(images=images)'.
Retrieve augmented labels via 'aug_det.augment_keypoints()' or 'aug_det.augment_bounding_boxes()'.
Visualize results using 'ia.imshow()' to verify data integrity before training.
All Set
Ready to go
Verified feedback from other users.
"Highly praised for its flexibility and ability to handle complex labels, though some users find the CPU-based performance slower than newer alternatives."
Post questions, share tips, and help other users.
Effortlessly find and manage open-source dependencies for your projects.

End-to-end typesafe APIs made easy.

Page speed monitoring with Lighthouse, focusing on user experience metrics and data visualization.

Topcoder is a pioneer in crowdsourcing, connecting businesses with a global talent network to solve technical challenges.

Explore millions of Discord Bots and Discord Apps.

Build internal tools 10x faster with an open-source low-code platform.

Open-source RAG evaluation tool for assessing accuracy, context quality, and latency of RAG systems.

AI-powered synthetic data generation for software and AI development, ensuring compliance and accelerating engineering velocity.