GreenMap
/

yolo11x-blueprint-legend-layout-detector

Object Detection

document-layout

Model card Files Files and versions

yolo11x-blueprint-legend-layout-detector / README.md

GreenMap's picture

update model

3deed2b 2 months ago

|

Raw History Blame

1.98 kB

	---
	license: apache-2.0
	library_name: ultralytics
	pipeline_tag: object-detection
	tags:
	- yolo
	- yolo11
	- object-detection
	- blueprint
	- floorplan
	- architecture
	- construction
	- legend
	- document-layout
	- cad
	---

	# GreenMap/yolo11x-blueprint-legend-detector

	![157](https://cdn-uploads.huggingface.co/production/uploads/689a00a12b3976126e5e8431/_HXzcYTKM6PsrvqRZswk5.jpeg)

	## Overview

	YOLO11x model trained to detect structural elements inside blueprint legends on architectural and construction drawings.

	- Detects legend symbols (hatch patterns and graphical symbols).
	- Detects legend descriptions.
	- Detects complete legend rows.
	- Detects legend continuation blocks.
	- Trained on architectural floor plans and construction drawings.

	## Classes

	\| ID \| Class \| Purpose \|
	\|----\|---------------------------\|---------------------------------------------------------------\|
	\| 0 \| legend_symbol \| Graphical symbol or hatch pattern representing a construction element or material. \|
	\| 1 \| legend_description \| Text description corresponding to a legend symbol. \|
	\| 2 \| legend_row \| Complete legend entry containing a symbol and its associated description. \|
	\| 3 \| legend_continuation_block \| A block containing the continuous legend elements. \|

	## Training Metrics

	![results](https://cdn-uploads.huggingface.co/production/uploads/689a00a12b3976126e5e8431/ltXn3VCWQaXr4nI1Xznry.png)

	## Validation Metrics

	\| Class \| Images \| Instances \| Precision \| Recall \| mAP50 \| mAP50-95 \|
	\|-------\|--------\|-----------\|-----------\|--------\|-------\|----------\|
	\| all \| 62 \| 2407 \| 0.896 \| 0.873 \| 0.895 \| 0.677 \|
	\| legend_symbol \| 62 \| 802 \| 0.977 \| 0.968 \| 0.975 \| 0.885 \|
	\| legend_description \| 62 \| 779 \| 0.966 \| 0.950 \| 0.973 \| 0.750 \|
	\| legend_row \| 62 \| 762 \| 0.927 \| 0.950 \| 0.967 \| 0.716 \|
	\| continuous_elements_block \| 44 \| 64 \| 0.714 \| 0.625 \| 0.665 \| 0.357 \|

	## Training

	- Base model: YOLO11x
	- Task: Object Detection
	- Input size: 1472 × 1472
	- Framework: Ultralytics YOLO