Skip to Content
Applied SciencesApplied Sciences
  • Article
  • Open Access

20 November 2026

30 Pages

PatMamba: Grid-Aware Vision Mamba for Compound Patent Figure Segmentation

and
Library, China University of Geosciences (Wuhan), Wuhan 430074, China
*
Author to whom correspondence should be addressed.
Appl. Sci.2026, 16(22), 11522;https://doi.org/10.3390/app162211522 
(registering DOI)
This article belongs to the Special Issue Advances in Computer Vision and Digital Image Processing

Abstract

Patent documents express much of their technical content in drawings, yet a typical compound figure packs several sub-figures together with a dense overlay of reference numerals and lead lines, so downstream tasks such as prior art search and figure-text association must first decompose the figure and separate the annotation text from the line-art. Existing methods treat this as a binary “sub-figure versus background” segmentation that discards the text as noise and encodes no structural prior. We propose PatMamba, a Vision Mamba framework built on a VMamba SS2D backbone and a U-Net decoder that jointly predicts three semantic classes, namely background, sub-figure regions and text-label regions. PatMamba also incorporates a grid-aware scan that appends two grid row/column-priority directions to the four cross-scan directions, thereby encoding the compound-figure layout into the selective-scan sequence at negligible parameter cost. On the DeepPatent2 benchmark, the Mamba backbone lifts figure-level sub-figure detection from 70.7% for the traditional run-length-smoothing baseline to 94.5% hit@IoU0.7, ahead of the CNN (72.6%) and Transformer (93.7%) backbones; the grid-aware scan lifts the pixel-level mIoU to 0.918 and yields its largest gains on dense compound figures with four or more sub-figures, improving the sub-figure IoU by +1.3 and +1.7 points and the Boundary F1 by +0.7 points. In zero-shot transfer to utility patents, biomedical compound figures, chemical reaction schemes and handwritten circuit diagrams, this structure-aware bias persists on the visually closest domain and decays in a regular gradient with domain distance.

Article Metrics

Citations

Article Access Statistics

Article metric data becomes available approximately 24 hours after publication online.