dask-image: distributed image processing for large data
summary
dask-image provides chunked lazy loading, ND filters, morphology and measurements that scale from laptop to cluster and can place chunk arrays onto the GPU via cuPy for acceleration. It composes pipelines with dask task graphs and exposes map_overlap/map_blocks for custom analysis and dask-image primitives for labeling and region measurements. A published benchmark on randomly generated 2 TB test: 2.5 h → <20 min with 8 GPUs demonstrates large speedups over single‑CPU processing.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:18Who needs daskimage
- 3:56Getting started
- 4:11Subpackages
- 6:07GPU support
- 8:18Building a pipeline
- 9:40Filtering
- 10:12Segmenting
- 11:17Local thresholding
- 12:41Morphological operations
- 18:11Measure labels
- 21:42Apply parallel
- 22:40Questions
related talks
01h 03mDask: Next Steps in Parallel Python
01h 23mData Science at Scale with Dask
01h 22mUsing Dask for Parallel Computing in Python
01h 26mDask tutorial
44mUsing Dask for Parallel Computing in Python
38mScalable Machine Learning with Dask
30mDask ❤️ Xarray: Geoscience at Massive Scale
02h 18mHacking Dask: Diving Into Dask;s Internals