Document Image Analysis Projects | OCR · Layout · Table · Form

Document Image Analysis Projects for Final Year Students (2026)

Document image analysis extracts text, structure and fields from scanned pages, photos of papers and PDFs. Student projects combine classical OpenCV preprocessing with OCR engines and modern layout models for tables, forms and multi-column pages.

This page lists 90+ high-impact topics. Tools include OpenCV, Tesseract, EasyOCR, PaddleOCR, Detectron2/YOLO for layout, LayoutLM/Donut and datasets such as PubLayNet, FUNSD and SROIE. Ideal for BE, BTech, MTech CS, ECE and AI students in Bangalore and across India.

Core Frameworks & Tools

Libraries and models commonly used in academic document image analysis projects.

OpenCV Tesseract EasyOCR / PaddleOCR LayoutLM / Donut Detectron2 / YOLO PubLayNet · FUNSD · SROIE

Best Document Image Analysis Topics & Tools (90+)

Grouped by theme. Each topic lists primary tools and typical datasets.

#	Project Topic	Tools · Datasets
🔤 OCR · Text Extraction
1	OCRTesseract OCR Pipeline for Scanned Pages	pytesseract, OpenCV
2	OCREasyOCR Multi-Language Text Extraction	EasyOCR
3	OCRPaddleOCR Detection + Recognition	PaddleOCR
4	OCRCompare Tesseract vs EasyOCR Accuracy	CER / WER metrics
5	OCRMobile Photo Document OCR with Deskew	Perspective + OCR
6	OCRLow-Quality / Noisy Scan OCR Robustness	Denoise + OCR
7	OCRMulti-Column Page OCR Reading Order	Column detect + OCR
8	OCRPDF Image Layer OCR Batch Pipeline	pdf2image + OCR
9	OCRWord-Level Bounding Boxes and Export	OCR boxes → JSON
10	OCRLanguage Detection before OCR	langdetect + engine
🧹 Preprocessing · Enhancement
11	PreDocument Skew Detection and Correction	Hough / projection
12	PreBinarization: Otsu, Adaptive, Sauvola	OpenCV thresholds
13	PrePerspective Correction of Phone Captures	Four-point transform
14	PreShadow and Illumination Normalization	CLAHE, morphology
15	PreNoise Removal and Speckle Cleaning	Morphology, median
16	PreBorder / Margin Detection and Crop	Contours
17	PreDPI and Resolution Standardization	Resize, pad
18	PreCompare Preprocessing Impact on OCR CER	Ablation table
📐 Layout Analysis · Segmentation
19	LayPage Segmentation: Text / Image / Table Regions	Contour / ML layout
20	LayPubLayNet-Style Layout Detection	Detectron2 / YOLO
21	LayReading Order Reconstruction	XY-cut / heuristics
22	LayHeader / Footer / Body Separation	Zone classification
23	LayMulti-Column Layout Detection	Projection profiles
24	LayFigure and Caption Association	Proximity + OCR
25	LayTitle and Heading Hierarchy Extraction	Font size / style
26	LayLayout Graph Representation of a Page	Blocks as nodes
📊 Table Detection · Extraction
27	TabTable Detection with Contours / Lines	OpenCV Hough lines
28	TabDeep Table Detection (YOLO / Detectron)	PubTables-style
29	TabCell Segmentation and Grid Reconstruction	Row/col lines
30	TabTable to CSV / Structured Export	OCR cells → CSV
31	TabMerged Cell and Complex Table Handling	Span detection
32	TabBorderless Table Detection Heuristics	Alignment clustering
33	TabFinancial Statement Table Extraction	Domain tables
34	TabCompare Classical vs Deep Table Models	IoU / structure F1
📝 Forms · Key-Value · Invoices
35	FormForm Field Detection and Labeling	FUNSD-style
36	FormKey–Value Pair Extraction from Forms	Layout + OCR
37	FormInvoice Field Extraction (date, amount, vendor)	SROIE / custom
38	FormReceipt Parsing Pipeline	OCR + rules / ML
39	FormCheckbox / Radio Button State Detection	Template match
40	FormSignature Region Detection	Ink density / CNN
41	FormID Card / KYC Field Extraction	Template + OCR
42	FormLayoutLM / Donut Document Understanding Demo	Transformers
✍️ Handwriting · HTR
43	HandHandwritten Digit Recognition (MNIST-style)	CNN, MNIST
44	HandOffline Handwritten Text Recognition Lite	CRNN / CTC concepts
45	HandSignature Verification (same / different)	Siamese / features
46	HandHandwritten Form Field OCR	Crop + HTR
47	HandLine Segmentation for Handwritten Pages	Projection / ML
48	HandCompare Printed vs Handwritten Regions	Classifier
🏷️ Document Classification · Type
49	ClsDocument Type Classification (invoice, ID, letter)	CNN / features
50	ClsPage Orientation Classification	0/90/180/270
51	ClsLanguage of Document Image	Script detection
52	ClsBlank vs Content Page Detection	Ink density
53	ClsQuality Score: Blur / Contrast Ranking	Laplacian, hist
54	ClsMulti-Label Document Attribute Tags	Has-table, has-form
🏢 Domain Applications
55	AppBank Cheque MICR / Amount Extraction	ROI + OCR
56	AppBusiness Card Digitization	Layout + OCR fields
57	AppMedical Prescription Digitisation Lite	Handwritten + print
58	AppExam Answer Sheet Region Mapping	Template zones
59	AppLegal Contract Clause Block Detection	Layout + headings
60	AppNewspaper Article Segmentation	Multi-column layout
61	AppHistorical Document Enhancement + OCR	Degrade reverse
62	AppShipping Label / Address Block Extraction	Region + OCR
📈 Evaluation · Metrics
63	EvalCharacter / Word Error Rate (CER / WER)	editdistance
64	EvalLayout Detection mAP / IoU Metrics	COCO-style eval
65	EvalTable Structure Recognition F1	Cell-level match
66	EvalEnd-to-End Field Extraction Accuracy	Key exact match
67	EvalHuman-in-the-Loop Correction Interface	Review UI
68	EvalBenchmark: 3 Engines × 3 Document Types	Unified report
🔬 Advanced · Multimodal · Deployment
69	AdvLayoutLM Fine-Tuning for Key Information	Transformers, FUNSD
70	AdvDonut End-to-End Document Parsing	Vision encoder-decoder
71	AdvMulti-Page Document Stitching and Index	Page order + OCR
72	AdvSynthetic Document Generation for Training	Augmentation pipeline
73	AdvActive Learning for Layout Annotation	Uncertainty sampling
74	AdvMobile Edge OCR with Quantized Models	TFLite / ONNX
75	AdvStreamlit Document Analysis Demo App	Upload → results
76	AdvFastAPI Document Processing Service	REST + async jobs
77	AdvSearchable PDF Generation from Scans	OCR layer embed
78	AdvPrivacy: Redact Sensitive Regions	Detect + blackout
79	AdvMultilingual Document Pipeline	Script + engine select
80	AdvCamera Document Capture Quality Gate	Blur / glare checks
81	AdvGraphical UI Element Detection in Docs	Icons / logos
82	AdvCross-Domain Transfer: Forms → Invoices	Fine-tune study
83	AdvExplainability: Why This Field Mapped	Attention / rules log
84	AdvBatch Processing Throughput Optimization	Parallel workers
85	AdvDataset Curation and Annotation Guidelines	Label schema
86	AdvFailure Mode Catalog: Skew, Blur, Bleed	Case study report
87	AdvEducational Lab: Preprocess → OCR → Layout → Export	Curriculum path
88	AdvOpen-Source Stack Comparison Report	Tesseract vs Paddle vs commercial
89	AdvReproducibility: Seeds, Configs, Version Pins	Environment lock
90	AdvEnd-to-End: Capture → Enhance → OCR → Structure → Export → Report	Full pipeline
91	AdvThesis Package: Methods, Metrics, Ablations, Discussion	Full documentation
92	AdvProduction Checklist: Latency, Accuracy, Languages, Formats	Ops report

Topics reflect document image analysis and OCR academic practice. Contact us for pipelines, evaluation metrics, university-format report, PPT and viva Q&A for any topic above.

Why Choose Us for Document Image Analysis Projects?

Bangalore-based guidance for BE, BTech and MTech students working on OCR, layout and form understanding systems.

OCR Pipelines

Tesseract, EasyOCR and PaddleOCR with preprocessing for real scans and photos.

Layout Analysis

Region detection, reading order and multi-column structure recovery.

Tables & Forms

Table structure extraction and key–value form field parsing.

Document AI

LayoutLM/Donut-style understanding and end-to-end field extraction.

Frequently Asked Questions — Document Image Analysis

Top topics include OCR with Tesseract/EasyOCR, document layout analysis, table structure extraction, form field detection, invoice/receipt parsing, handwritten recognition and LayoutLM-style document understanding.

OpenCV, Tesseract OCR, EasyOCR, PaddleOCR, Detectron2/YOLO for layout, LayoutLM/Donut, scikit-image, and datasets such as PubLayNet, FUNSD, SROIE and custom scanned documents.

Yes. Packages include preprocessing pipelines, OCR/layout models, evaluation metrics, university-format report, PPT and viva Q&A.

It converts scanned or photographed documents into structured data — text, tables, form fields and layout regions — for digitization, archival search, invoice automation and accessibility.

Document Image Lab — Bangalore

OCR, layout, table and form setups for BE, BTech and MTech document AI projects.

OCR
Tesseract / EasyOCR

Text Extract

Layout
Segmentation

Regions

Table
Extraction

Structure → CSV

Form / Invoice
Fields

Key–Value

Handwriting
HTR

CRNN Concepts

Preprocessing
Skew / Binarize

OpenCV

Document AI
LayoutLM

Understanding

Viva & Report
Preparation

Documentation

Document Image Analysis Projects.