The description and classification of the ocrd-tool.json could better reflect what is really going on:
description = 'Detect lines' – please state here what the README already says
steps = 'layout/segmentation/line' – should also have layout/segmentation/region (or just layout/segmentation), maybe even preprocessing/optimization/deskewing (IIUC)
categories = "Image preprocessing", "Layout analysis", "Text recognition and optimization" – should only be Layout analysis
mikegerber
commented
4 years ago
The description and classification of the
ocrd-tool.jsoncould better reflect what is really going on:description = 'Detect lines'– please state here what the README already sayssteps = 'layout/segmentation/line'– should also havelayout/segmentation/region(or justlayout/segmentation), maybe evenpreprocessing/optimization/deskewing(IIUC)categories = "Image preprocessing", "Layout analysis", "Text recognition and optimization"– should only beLayout analysis