光学字符识别 Jobs

there are total 235 questions. Need to extract only questions in word doc. No need to type answer or answer keys. I have sample format that must be followed

I have mapped out a mobile MVP that turns any high-school Biology or Chemistry syllabus into a personalised, evidence-based study schedule. A student snaps a photo of the hand-out or drops in the original PDF—both formats must work seamlessly—then the app extracts the key learning objectives, test dates and other cues and converts them into a day-by-day plan laid out as an intuitive visual timeline. Core flow 1. Scan / upload → run on-device OCR (Tesseract, VisionKit or similar). 2. Parse text with lightweight NLP to isolate learning objectives first, then dates, topics and assessments. 3. Feed the parsed data into a scheduling engine that applies retrieval practice, spacing and interleaving rules. 4. Serve the output as swipeable, colour-coded timeline cards (Rea...

**PROJECT DOCUMENT: FREELANCER REQUIREMENT – ANPR SDK DEVELOPMENT** Development of Offline ANPR (Automatic Number Plate Recognition) SDK Project Objective: To develop a high-performance, offline ANPR SDK capable of real-time license plate detection and recognition from IP camera streams, optimized for Indian road conditions. **Project Scope:** The selected freelancer/team will be responsible for designing and developing a modular ANPR SDK that can be integrated into edge devices or local servers. The SDK must process live RTSP streams and provide structured outputs via APIs. **Functional Requirements:** 1. Real-time video stream processing (RTSP / IP cameras) 2. Vehicle and number plate detection 3. Character segmentation and OCR 4. Output structured data: * Vehicle numbe...

OCR System for Live Poker Video — Master's Research Project I'm looking for an experienced developer to build a precise OCR/computer vision module for a master's research project. The system reads game data (card ranks, suits, player stacks, pot size) from a 1080p poker client video feed. Current setup: Video input: 1080p, 30fps Platform: Windows, Python What I need: High accuracy recognition of card ranks (A K Q J T 9–2) and suits on fixed screen regions Robust numeric OCR for stack/pot amounts (variable text) Must handle compression artifacts typical of video feeds. must handle normalization for more than 1 table analisis. In your proposal, please include: Final price for the complete module Which methods/approaches you would use to achieve the required ...

I have a set of PDFs that contain purely textual information presented in a mix of paragraphs, scattered labels, and occasional row-style groupings. I need every single data point moved into a clean Excel workbook so nothing is lost in translation. Because the layout switches between narrative blocks and ad-hoc table-like sections, the job will likely involve both automated capture (Python, Power Query, or your preferred OCR tool) and some careful manual cleanup to preserve order and context. Deliverable: one well-structured .xlsx file per source PDF, with all columns and data fields represented exactly as they appear, ready for filtering and analysis. Consistent header naming and no stray line breaks or merged cells, please. If you’ve handled mixed-format PDFs before and can turn ...

I have a collection of mixed-content PDFs—some pages are pure scans, the rest contain searchable text. Each file holds key tables I need converted into clean, structured data and then visualised as clear bar charts for quick comparison. You will: • Isolate every table across the PDFs, using OCR where the page is only an image. • Clean and normalise the numbers so columns and units stay consistent. • Produce bar charts that faithfully reflect the extracted figures (one chart per table unless otherwise noted). Deliverables 1. A single, tidy dataset (CSV or Excel) with each table clearly identified. 2. High-resolution bar chart images (PNG or SVG) and the source file (Excel, Power BI, or Python notebook) so I can regenerate them later. 3. A short note outlinin...

Engenheiro de IA Sênior - Otimização de Pipeline de Longo Contexto (Python) Descrição do Projeto: Estamos buscando um especialista em IA para destravar e otimizar nosso pipeline de análise de matrículas imobiliárias. O sistema é construído em Python e utiliza o Gemini 1.5 Pro/Flash para processar documentos que podem chegar a centenas de páginas. O foco é garantir que o modelo processe o contexto integral sem truncamento e com rigor lógico. Principais Desafios e Responsabilidades: * Sanitização e Ingestão: Refinar a extração de texto (OCR/Parsing) para evitar estouro de buffers e garantir que o texto completo chegue à IA via Python/MySQL (uso de LONGTEXT). * G...

I have a batch of PDFs made up entirely of scanned images. The scans are of medium quality, so the text is readable but not crystal-clear; careful OCR work and manual proofreading will be required. All I need back is clean, plain UTF-8 text—no layout reconstruction, styling, or headers necessary. Speed is important: this project is time-sensitive and I’m ready to start as soon as I select the freelancer. Tools such as Tesseract, ABBYY FineReader, or Adobe Acrobat’s OCR module are fine, provided the final output meets the accuracy standard below. Deliverables • A separate .txt file for each source PDF, named to match the original file • Spelling and obvious recognition errors corrected • Any unreadable characters clearly flagged with “[unreadable]&rd...

I have 50 high-resolution JPG/PNG pages that record Viennese street name changes—they map the historic to current ones. I need these pages converted into editable text with publication-grade accuracy. Because the files are already digital, no physical scanning is required; the job is all about precise OCR and meticulous proofreading. You may use any robust engine you prefer—ABBYY FineReader, a finely-tuned Tesseract model, or similar—but the end result must be spotless German text, including every umlaut and ß exactly as printed. The layout is simple: two columns per entry, so please preserve that structure. Add any additional information and problems you encounter. Deliverables • A UTF-8 Excel or CSV file with columns: old_name | new_name | page_number &b...

I have a batch of PDF files that hold pure text data organised in tables. Every cell must end up in the correct row and column inside Excel, no omissions, no merged-cell shortcuts. Some of the tables share a common layout, while others change their order of columns or include extra rows, so a simple copy-paste macro will not be enough—you will need to review each file and adapt where the structure shifts. Speed is important, yet not at the expense of accuracy; I expect every spreadsheet to match its source 100 %. Feel free to use Adobe Acrobat, Able2Extract, Power Query, or any OCR workflow you trust, as long as the final .xlsx files are clean and fully editable. Deliverables: • One Excel sheet (or workbook) per PDF, formatted in true tabular form • Headings preserved e...

I have 50 high-resolution JPG/PNG pages that record Viennese street name changes—they map the historic to current ones. I need these pages converted into editable text with publication-grade accuracy. Because the files are already digital, no physical scanning is required; the job is all about precise OCR and meticulous proofreading. You may use any robust engine you prefer—ABBYY FineReader, a finely-tuned Tesseract model, or similar—but the end result must be spotless German text, including every umlaut and ß exactly as printed. The layout is simple: two columns per entry, so please preserve that structure. Add any additional information and problems you encounter. Deliverables • A UTF-8 Excel or CSV file with columns: old_name | new_name | page_number &b...

I have a batch of PDF files that hold pure text data organised in tables. Every cell must end up in the correct row and column inside Excel, no omissions, no merged-cell shortcuts. Some of the tables share a common layout, while others change their order of columns or include extra rows, so a simple copy-paste macro will not be enough—you will need to review each file and adapt where the structure shifts. Speed is important, yet not at the expense of accuracy; I expect every spreadsheet to match its source 100 %. Feel free to use Adobe Acrobat, Able2Extract, Power Query, or any OCR workflow you trust, as long as the final .xlsx files are clean and fully editable. Deliverables: • One Excel sheet (or workbook) per PDF, formatted in true tabular form • Headings preserved e...

I’m extending our Zoho Creator app with two tightly-linked features and want a developer who already speaks the Zoho ecosystem fluently. 1. Conversational AI agent The first module should let clients type a question inside our existing Creator portal and instantly receive answers drawn from the invoice records already stored in that same Creator database. Think of it as a smart layer over Deluge queries: the agent interprets the user’s request, fetches the right data, formats a clear reply, then logs the entire exchange for audit. An OpenAI or similar LLM endpoint is fine as long as it sits securely behind Creator’s functions. 2. OCR-driven invoice capture Most invoices reach us as scanned paper PDFs or images. I need an OCR pipeline (Zoho’s own Vision, Google...

I have a batch of PDFs—some are pure digital files, others are scanned images—and I need to pull specific text fields from each of them into a clean, well-structured Excel workbook. Because this is a one-time job, I’m aiming for the fastest reliable approach rather than an enterprise-level pipeline. Scope • Identify and capture only the targeted text values I will specify (invoice numbers, dates, totals, etc.). • Handle both searchable PDFs and image-based pages in the same run, applying accurate OCR where needed. • Output every record in an .xlsx file with clearly labeled columns ready for further analysis. Preferred Stack Python with libraries such as pdfplumber, PyPDF2, or camelot for the text-based files, combined with Tesseract (or a comparab...

Looking for an experienced Android Java developer to build a complete app for an AR glasses device. The device connects to an Android phone via USB-C. The app must: — Open a live USB UVC camera feed fullscreen on the glasses display — Apply real-time zoom from 1x to 10x with sharpness maintained — Apply real-time contrast enhancement — Apply real-time brightness adjustment — Apply edge detection outline mode — Apply colour filter modes — Read visible text aloud using OCR and text to speech — Simple on-screen menu to switch between features — Maintain minimum 30 FPS at all times SDKs provided: — — USB camera, zoom, image effects (Java API over C/C++ native libraries) — — glasses communication and display cont...

i need to read only questions and type them exactly as they are in word doc. I don't want answers or answer keys

I will send you a list of five Kindle titles. Your task is to acquire each book from the Kindle Store, remove any DRM if necessary, and export a clean PDF that mirrors the exact Kindle layout—fonts, images, page breaks, everything must look just as it does on the device. Please enable OCR so every page is fully searchable. A generated Table of Contents should also be embedded in each file for easy navigation. I don’t need bookmarks or annotations beyond that. Use whichever tools you prefer—Calibre with the Kindle De-DRM plugin, Kindle for PC/Mac, Adobe Acrobat, or other reliable software—as long as the final PDFs open flawlessly on any reader and pass a quick text-search check. Deliverables (one set per book) • PDF replicating the Kindle layout • Se...