PNEngineeringDatasets commited on
Commit
620ab24
·
verified ·
1 Parent(s): 7736ca3

Upload 10 files

Browse files
DATASET_INFO_TEMPLATE.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_name": "RC_Walls_V1",
3
+ "dataset_version": "V1",
4
+ "element_type": "Reinforced Concrete Walls",
5
+ "pdf_count": 25,
6
+ "png_count": 25,
7
+ "description": "High-resolution RC column drawings suitable for AI training, OCR and education.",
8
+ "file_format_pdf": "PDF (flattened, metadata removed)",
9
+ "file_format_png": "PNG 1200 DPI (clean, high-resolution)",
10
+ "license": "CEDL",
11
+ "created_by": "PN Engineering Datasets"
12
+ },
13
+ "cleaning_process": [
14
+ "anonymized",
15
+ "metadata removed",
16
+ "flattened",
17
+ "noise reduction",
18
+ "background cleanup"
19
+ ],
20
+ "ai_use_cases": [
21
+ "OCR",
22
+ "Vision models",
23
+ "BIM automation",
24
+ "CAD detection",
25
+ "LLM multimodal training"
26
+ ],
27
+ "creator": "PN Engineering Datasets",
28
+ "year": 2025,
29
+ "license": "CEDL Extended"
30
+ }
DATASET_STRUCTURE_BLUEPRINT.txt ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ RC_Columns_V1/
2
+ ├── 01_PDF (25 files)
3
+ ├── 02_PNG (25 files)
4
+ ├── 03_Documentation/
5
+ ├── DATA_DICTIONARY
6
+ ├── DATA_INFO_TEMPLATE
7
+ ├── DATASET_STRUCTURE_BLUEPRINT
8
+ ├── index.csv
9
+ ├── LICENSE
10
+ ├── README
DATA_DICTIONARY.txt ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ PN Engineering Datasets
2
+ DATA DICTIONARY – Professional Specification
3
+ --------------------------------------------------
4
+
5
+ This document defines the structure and attributes of all datasets released
6
+ under PN Engineering Datasets.
7
+
8
+ --------------------------------------------------
9
+ 1. Dataset Metadata
10
+ --------------------------------------------------
11
+ dataset_name: RC_Walls_Dataset_V1
12
+ dataset_version: V1
13
+ element_type: Reinforced concrete columns
14
+ pdf_count: 25
15
+ png_count: 25
16
+
17
+ --------------------------------------------------
18
+ 2. File Structure
19
+ --------------------------------------------------
20
+ PDF files:
21
+ • Flattened
22
+ • Anonymous
23
+ • Metadata removed
24
+ • OCR-ready
25
+
26
+ PNG files:
27
+ • 1200 DPI resolution
28
+ • Clean and uniform background
29
+ • High contrast for Vision models
30
+
31
+ Documentation files:
32
+ • DATA_DICTIONARY
33
+ • DATASET_INFO_TEMPLATE.json
34
+ • DATASET_STRUCTURE_BLUEPRINT
35
+ • INDEX.csv
36
+ • LICENSE.txt
37
+ • README
38
+
39
+ --------------------------------------------------
40
+ 3. Drawing Contents
41
+ --------------------------------------------------
42
+ Each drawing may include:
43
+ • Reinforcement layouts
44
+ • Longitudinal bars
45
+ • Stirrups/ties
46
+ • Sections in 1–3 views
47
+ • Dimensions (millimeters)
48
+ • Structural symbols
49
+ • Work stages
50
+ • Concrete classes
51
+ • Notes and detailing conventions
52
+
53
+ --------------------------------------------------
54
+ 4. Data Preparation Process
55
+ --------------------------------------------------
56
+ • Anonymization (project & company data removed)
57
+ • Noise reduction
58
+ • PDF flattening
59
+ • Cleanup of backgrounds
60
+ • Export to high-resolution PNG
61
+
62
+ --------------------------------------------------
63
+ 5. Machine Learning Use-Cases
64
+ --------------------------------------------------
65
+ • OCR model training
66
+ • Vision model detection of bars, symbols, dimensions
67
+ • Generating embeddings for engineering LLMs
68
+ • CAD-to-AI pipelines
69
+ • BIM automation
70
+
71
+ --------------------------------------------------
72
+ 6. Limitations
73
+ --------------------------------------------------
74
+ Not for construction use.
75
+ For training & research only.
76
+
77
+ ----------------------------------------------
78
+ 7. SUPPORT
79
+ ----------------------------------------------
80
+
81
+ If you have questions, need additional formats, or want a custom dataset
82
+ (beams, slabs, foundations, walls, steel details, construction photos, BIM sets),
83
+ please contact:
84
+
85
+ PN Engineering Datasets
86
+ Email: pawel.niczyporuk@gmail.com
87
+
88
+
89
+
90
+ Thank you for your purchase!
91
+ PN Engineering Datasets – 2025
LICENSE.txt ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ PN Engineering Datasets
2
+ CEDL – Commercial Engineering Dataset License (Extended)
3
+
4
+ This license applies to:
5
+ RC_Walls_Dataset_V1
6
+ 25 PNG drawings + 25 PDF drawings
7
+ © PN Engineering Datasets
8
+
9
+ --------------------------------------------------
10
+ 1. Permissions
11
+ --------------------------------------------------
12
+ You MAY:
13
+ - Use the dataset for commercial or non-commercial purposes.
14
+ - Train AI models, machine learning systems, OCR tools, BIM tools,
15
+ or any algorithm on these drawings.
16
+ - Use the dataset internally within your company or organization.
17
+ - Include results, metrics, screenshots, derivations, extracted data,
18
+ or model outputs in commercial projects.
19
+ - Transform, annotate, or preprocess the dataset for your own workflows.
20
+
21
+ --------------------------------------------------
22
+ 2. Restrictions
23
+ --------------------------------------------------
24
+ You MAY NOT:
25
+ - Resell, redistribute, upload publicly, or give away the raw dataset
26
+ (PNG/PDF files) in its original or slightly modified form.
27
+ - Include the dataset in any free or paid public repository (e.g., Kaggle,
28
+ HuggingFace, GitHub) unless it is significantly transformed
29
+ and cannot be reconstructed.
30
+ - Claim authorship of the original drawings.
31
+
32
+ --------------------------------------------------
33
+ 3. Allowed Public Outputs
34
+ --------------------------------------------------
35
+ You MAY publish:
36
+ • AI models trained on the dataset
37
+ • Annotations, masks, bounding boxes
38
+ • OCR outputs
39
+ • Vectorizations
40
+
41
+ 4. No Warranty
42
+ --------------------------------------------------
43
+ Provided “as-is”. Not suitable for real-world construction use.
44
+
45
+ ------------------------------------------------
46
+ 5. Contact
47
+ ------------------------------------------------
48
+ For commercial licensing, custom datasets, or bulk data acquisition:
49
+ PN Engineering Datasets
50
+ [pawel.niczyporuk@gmail.com]
51
+
52
+
53
+ PN Engineering Datasets – 2025
RC_Wall_V1-W005-2414_1.png ADDED

Git LFS Details

  • SHA256: 2ab7034d41b492b80e23a64212fe27d8cce41dc8fd4b3b2b5900cdff4213c854
  • Pointer size: 132 Bytes
  • Size of remote file: 7.78 MB
RC_Wall_V1-W010-2414_1.png ADDED

Git LFS Details

  • SHA256: 40cc1f630070106411197139d7cdf1bcb8b3896a767ef30ed315f836b8a3f2e8
  • Pointer size: 131 Bytes
  • Size of remote file: 379 kB
RC_Wall_V1-W019-2414_1.png ADDED

Git LFS Details

  • SHA256: 2adef2c33278526a2ab90144a812e0680ac6139c83300201055123bed40ca478
  • Pointer size: 132 Bytes
  • Size of remote file: 2.36 MB
RC_Wall_V1-W020-2414_1.png ADDED

Git LFS Details

  • SHA256: 859ffd2eb7de58f3981241840ba51b72969f98a3db9982b03dbd712b7efa34a1
  • Pointer size: 132 Bytes
  • Size of remote file: 1.98 MB
RC_Wall_V1-W025-2414_1.png ADDED

Git LFS Details

  • SHA256: a56084ebbf9dbd86ba6d36427fc012f37ef10dfec8df0d5058275761b0e637dd
  • Pointer size: 132 Bytes
  • Size of remote file: 1.86 MB
README.txt ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ PN Engineering Datasets
2
+ README – Professional Documentation
3
+ --------------------------------------------------
4
+
5
+ Thank you for using PN Engineering Datasets.
6
+ This dataset contains real reinforced concrete (RC) engineering drawings
7
+ created by a professional structural engineering office with 15+ years of experience.
8
+
9
+ --------------------------------------------------
10
+ 1. Purpose of the Dataset
11
+ --------------------------------------------------
12
+ The dataset is designed for:
13
+ • AI/ML training (OCR, Vision, multimodal LLMs)
14
+ • Structural engineering education
15
+ • BIM automation development
16
+ • CAD/drafting learning
17
+ • Civil engineering research
18
+ • Dataset benchmarking
19
+
20
+ --------------------------------------------------
21
+ 2. Dataset Structure
22
+ --------------------------------------------------
23
+ /01_PDF/
24
+ Cleaned, flattened, anonymized PDF drawings
25
+
26
+ /02_PNG/
27
+ High-resolution PNG (1200 DPI) versions
28
+
29
+ /03_Documentation/
30
+ DATA_DICTIONARY, DATASET_INFO_TEMPLATE, DATASET_STRUCTURE_BLUEPRINT, JSON metadata, License, README
31
+
32
+ --------------------------------------------------
33
+ 3. File Formats
34
+ --------------------------------------------------
35
+ PDF:
36
+ • Flattened
37
+ • Metadata removed
38
+ • OCR-friendly
39
+
40
+ PNG:
41
+ • 1200 DPI
42
+ • Clean background
43
+ • High contrast
44
+
45
+ --------------------------------------------------
46
+ 4. Usage Guidelines
47
+ --------------------------------------------------
48
+ You MAY:
49
+ • Use for internal commercial AI training
50
+ • Train OCR/CV/LLM models
51
+ • Use in academic or enterprise research
52
+ • Build derived datasets/models
53
+
54
+ You MAY NOT:
55
+ • Redistribute or resell raw files
56
+ • Upload raw drawings publicly
57
+ • Claim authorship of the original drawings
58
+
59
+ This dataset is suitable for commercial use inside your own tools,
60
+ models, or research workflows. Redistribution of the dataset itself
61
+ is not permitted — see LICENSE.txt for full details.
62
+
63
+ --------------------------------------------------
64
+ 5. Software Compatibility
65
+ --------------------------------------------------
66
+ The dataset works with:
67
+ • Python, OpenCV, Tesseract, PIL
68
+ • PyTorch / TensorFlow
69
+ • Roboflow
70
+ • All major OCR/CV frameworks
71
+ • Engineering & BIM automation tools
72
+
73
+ --------------------------------------------------
74
+ 6. Notes
75
+ --------------------------------------------------
76
+ Not intended for real construction use.
77
+ Provided strictly for research and training purposes.
78
+
79
+ ----------------------------------------------
80
+ 7. SUPPORT
81
+ ----------------------------------------------
82
+
83
+ If you have questions, need additional formats, or want a custom dataset
84
+ (beams, slabs, foundations, walls, steel details, construction photos, BIM sets),
85
+ please contact:
86
+
87
+ PN Engineering Datasets
88
+ Email: pawel.niczyporuk@gmail.com
89
+
90
+
91
+
92
+ Thank you for your purchase!
93
+ PN Engineering Datasets – 2025