Parallel watershed partitioning: GPU-based hierarchical image segmentation

• Published in: Journal of parallel and distributed computing Vol. 205; p. 105140
• Main Authors: Yeghiazaryan, Varduhi, Gabrielyan, Yeva, Voiculescu, Irina
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.11.2025
• Subjects: GPU acceleration; Parallel union–find; Path reduction; Real-time image segmentation; Waterfall transform; Watershed transform; Path reduction; Real-time image segmentation; Watershed transform; Waterfall transform; Parallel union–find; GPU acceleration
• ISSN: 0743-7315
• DOI: 10.1016/j.jpdc.2025.105140

Many image processing applications rely on partitioning an image into disjoint regions whose pixels are ‘similar.’ The watershed and waterfall transforms are established mathematical morphology pixel clustering techniques. They are both relevant to modern applications where groups of pixels are to be decided upon in one go, or where adjacency information is relevant. We introduce three new parallel partitioning algorithms for GPUs. By repeatedly applying watershed algorithms, we produce waterfall results which form a hierarchy of partition regions over an input image. Our watershed algorithms attain competitive execution times in both 2D and 3D, processing an 800 megavoxel image in less than 1.4 sec. We also show how to use this fully deterministic image partitioning as a pre-processing step to machine-learning-based semantic segmentation. This replaces the role of superpixel algorithms, and results in comparable accuracy and faster training times. The code is publicly available at https://github.com/hamemm/PRUF-watershed.git. •Three parallel watershed algorithms for execution on GPUs, offering significant speedup over existing approaches.•Introduction of intermediate GPU algorithmic steps to generate hierarchical segmentations through the waterfall transform.•Applications of the proposed approaches in deep-learning pipelines, contributing to speedup and/or accuracy improvement.