ilastik review
An open-source desktop toolkit for annotation-driven classification and scientific image analysis.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
For scientific image analysis, ilastik brings segmentation, classification, tracking, and counting workflows into a downloadable desktop toolkit. Its central pixel-classification workflow learns from user-drawn annotations with a Random Forest classifier, then produces semantic class predictions. Researchers who want to interactively label pixels and analyze scientific images are its clearest fit; the software is available for Windows, macOS, and Linux and is published under the GNU General Public License.
The workflow set extends beyond pixel classification. Semantic segmentation can produce probability maps and class-label exports, while Autocontext adds cascaded pixel classification. For other tasks, ilastik includes object classification based on object-level features and annotations, boundary-based segmentation with Multicut, and semi-automatic 2D and 3D carving. Tracking workflows cover objects in 2D+time and 3D+time datasets. Together, these features make the toolkit useful across several stages of scientific image analysis, though its approach remains annotation-driven rather than a general-purpose segmentation service. Temporal video data is handled through tracking workflows, not as a standalone video-segmentation product; instance segmentation is not supported.
ilastik is open source and free, with desktop and self-hosted deployment options rather than a subscription-plan structure. Trained classifiers can be applied to additional images through batch and headless processing, which suits repeatable processing without requiring each image to be annotated anew. Fiji is an integration, and community support and documentation are the listed support channels. Researchers seeking interactive, local analysis with pixel classification, tracking, and batch processing should consider ilastik. Teams whose requirements center on instance segmentation or a standalone video-segmentation product should look for a tool built around those needs instead.
ilastik pros and cons
- Where it wins
- Trains Random Forest pixel classifiers from user-drawn annotations
- Includes semantic segmentation, carving, object classification, and tracking workflows
- Runs trained classifiers in batch or headless mode
- Where it doesn't
- Pixel classification depends on user-drawn annotations
- Instance segmentation is not supported
- Temporal segmentation is part of tracking workflows, not a standalone video tool
ilastik fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. We may earn a commission when you click some links. How we rank.
Last updated · How we research and update