alibaba-quark-scanking

# Tool List > **All-in-one AI Document Scanning and Processing Open Platform** — 36 AI capabilities, 5 integration methods, covering text recognition, document scanning, format conversion, and AIGC. --- <p>Official Website: https://scan.quark.cn/business</p> <p>Cooperation Hotline: 4009030958</p> <p>Cooperation Email: quark-ai@service.alibaba.com</p> --- <p> <img alt="License" src="https://img.shields.io/badge/license-Apache--2.0-blue.svg"> <img alt="Python" src="https://img.shields.io/badge/python-3.9%2B-green.svg"> <img alt="OpenClaw" src="https://img.shields.io/badge/OpenClaw-Compatible-orange.svg"> <img alt="Platform" src="https://img.shields.io/badge/platform-macOS%20%7C%20Linux%20%7C%20Windows-lightgrey.svg"> <img alt="Scenes" src="https://img.shields.io/badge/scenes-30%2B-brightgreen.svg"> <img alt="Stars" src="https://img.shields.io/github/stars/YOUR_ORG/openclaw-yescan-skills?style=social"> </p> --- ## Table of Contents - [Platform Introduction](#平台简介) - [Core Advantages](#核心优势) - [Capability Overview](#能力全景) - [Text Recognition OCR（19 items）](#文字识别-ocr19项) - [Document Scanning Scan（13 items）](#文件扫描-scan13项) - [Format Conversion Typeset（3 items）](#格式转换-typeset3项) - [AI ID Photo AIGC（1 item）](#ai-证件照-aigc1项) - [Five Integration Methods](#五种接入方式) - [User Scenarios](#用户场景) - [Customer Cases](#客户案例) - [Billing Scheme](#计费方案) - [Quick Start](#快速开始) - [Frequently Asked Questions](#常见问题) - [Contact Us](#联系我们) --- ## Platform Introduction Quark Scan Open Platform is an AI document processing PaaS platform under Alibaba's Intelligent Information Business Unit, originating from Quark Scan App's core AI technology precipitation for hundreds of millions of users. The platform provides **36 AI capabilities** for B-end developers and enterprises, covering text recognition (OCR), document scanning enhancement, format conversion, and AI ID photo generation, supporting **20+ languages** and handling millions of documents daily. **One API, 36 capabilities** — all capabilities are accessed through a unified Vision API endpoint, allowing developers to switch parameters to call different capabilities, greatly reducing integration costs. --- ## Core Advantages ### 🎯 **High-Precision Text Recognition** Based on deep learning OCR engine, accuracy is at the top level of the industry, accurately recognizing printed, handwritten, mathematical formulas, and multi-language mixed typesetting. ### 📐 **Intelligent Typesetting Restoration** Large models deeply understand document layout structures, tables, and text mixed typesetting can be accurately restored, and output maintains the original layout of Word/Excel/PDF documents. ### 🔍 **Extremely High-Quality Scanning** Ultra-clear scanning quality, intelligent identification and erasure of handwritten words, watermarks, shadows, screen texture, and other impurities, restoring clean and clear document images. ### 🔌 **Five Integration Methods** API / MCP / OpenClaw / CLI / Hermes Agent — covering direct integration, AI Agent collaboration, command-line scripts, and other full-scenarios, with 1-minute integration. ### 🛡️ **Enterprise-Level Security** SHA3-256 signature authentication + multi-layer current limiting (global QPS / account QPS / capability frequency) + data non-disk storage options, meeting enterprise-level security compliance requirements. --- ## Capability Overview ``` ┌──────────────────────────────────────────────────────────────────────┐ │ Quark Scan Open Platform · 36 AI Capabilities │ ├─────────────┬────────────────────────────────────────────────────────┤ │ Text Recognition OCR │ 19 items: general text / handwritten text / table / ID card / │ │ (19 items) │ business license / social security card / driver's license / │ │ │ vehicle registration / VAT invoice / train ticket / │ │ │ Hong Kong and Macau travel permit / degree certificate / formula / │ │ │ question / medical report / drug report / commodity picture / │ │ │ picture translation │ ├─────────────┼────────────────────────────────────────────────────────┤ │ Document Scanning │ 13 items: scanned file / test paper enhancement / ID card / │ │ (13 items) │ document enhancement / remove handwriting / remove watermark / │ │ │ remove shadow / remove screen texture / image quality enhancement / │ │ │ document background removal / contract organization / sketch / │ │ │ line draft extraction / image cropping and correction │ ├─────────────┼────────────────────────────────────────────────────────┤ │ Format Conversion │ 3 items: image to Word / image to Excel / image to PDF │ │ (3 items) │ │ ├─────────────┼────────────────────────────────────────────────────────┤ │ AI ID Photo │ 1 item: intelligent ID photo generation (9 styles / 3 color backgrounds / clarity adjustment) │ │ (1 item) │ │ └─────────────┴────────────────────────────────────────────────────────┘ ``` ### Text Recognition OCR（19 items） Full-coverage recognition capabilities from general text to vertical scenarios, supporting printed, handwritten, and structured field extraction. | Capability | Description | Applicable Scenarios | |------|------|---------| | **General Text Recognition** | Supports 20+ languages, returns text content and coordinates | Document digitalization, content extraction | | **Handwritten Text Recognition** | High-precision recognition of handwritten text | Handwritten notes, annotation recognition | | **Table Recognition** | Structured table extraction, including row and column merging | Financial statements, data entry | | **ID Card Recognition** | Full-field structured extraction of front and back | Real-name authentication, account opening | | **Business License Recognition** | 14 field structured extraction | Enterprise access review | | **VAT Invoice** | Full information extraction of invoice | Reimbursement, financial system | | **Driver's License / Vehicle Registration** | Key information structured extraction | Vehicle management, insurance | | **Social Security Card / Hong Kong and Macau Travel Permit / Degree Certificate** | Structured recognition of certificates | Government, personnel management | | **Train Ticket / Overseas Ticket** | Ticket information extraction | Travel reimbursement | | **Formula Recognition** | LaTeX formula output | Education, academic | | **Question Recognition** | Structured splitting of test questions | Education technology, online learning | | **Medical Report / Drug Report** | Professional medical document recognition | Medical informatization | | **Commodity Picture Recognition** | Commodity label + barcode recognition | E-commerce, retail | | **Picture Translation** | Translate text in pictures and embed in original picture | Cross-language document processing | ### Document Scanning Scan（13 items） AI-driven document image enhancement and processing, restoring mobile phone-scanned documents to scanner-level quality. | Capability | Description | Applicable Scenarios | |------|------|---------| | **Scanned File** | Intelligent cropping + correction + enhancement, one-click generation of scanned file | Document archiving, digitalization | | **Test Paper Enhancement** | Test paper-specific optimization, improving clarity | Education scenarios | | **ID Card / Document Enhancement** | Quality optimization of ID card pictures | Identity authentication | | **Remove Handwriting** | Intelligent erasure of handwritten words, retaining printed content | Test paper correction, document sorting | | **Remove Watermark** | Remove document watermark | Document restoration | | **Remove Shadow** | Eliminate photographing shadows | Mobile phone document scanning | | **Remove Screen Texture** | Eliminate screen flipping's Moiré pattern | Screen flipping | | **Image Quality Enhancement** | Low-resolution image super-clear enhancement | Old documents | | **Document Background Removal** | Remove colored background and convert to pure white | Document printing | | **Contract Organization** | Contract document-specific organization and enhancement | Legal scenarios | | **Sketch** | Photo to sketch style | Creative applications | | **Line Draft Extraction** | Extract image outline lines | Design reference | | **Image Cropping and Correction** | Automatic detection of document boundary cropping | Document scanning pre-processing | ### Format Conversion Typeset（3 items） Intelligent conversion of image-level documents to editable files, maintaining original layout. | Capability | Description | Output Format | |------|------|---------| | **Image to Word** | Retain original layout, generate Word document | .docx | | **Image to Excel** | Table picture restoration to editable table | .xlsx | | **Image to PDF** | Image document conversion to searchable PDF | .pdf | ### AI ID Photo AIGC（1 item） Large model-driven intelligent ID photo generation, supporting rich customization options. | Function | Options | |------|------| | **Style** | Original, delicate, rosy, matte, white, skin, child, Korean, natural (9 styles) | | **Background Color** | Red, white, blue | | **Clarity** | 4-level adjustable (0-3) | | **Applicable Scenarios** | Employee entry, exam registration, social security, fund, visa application, resume creation | --- ## Five Integration Methods ### Method 1: **Direct API Integration (Recommended)** The most standard integration method, suitable for teams with backend development capabilities to directly embed into business systems. ``` POST https://scan-business.quark.cn/vision Content-Type: application/json ``` **Python Code Example**: ```python from time import time import requests, uuid, hashlib client_id = "your-client-id" client_secret = "your-client-secret" sign_nonce = uuid.uuid4().hex timestamp = int(time() * 1000) # SHA3-256 signature raw = f"{client_id}_vision_SHA3-256_{sign_nonce}_{timestamp}_{client_secret}" signature = hashlib.sha3_256(raw.encode("utf-8")).hexdigest() resp = requests.post("https://scan-business.quark.cn/vision", json={ "dataUrl": "https://example.com/document.jpg", "dataType": "image", "serviceOption": "ocr", "inputConfigs": '{"function_option":"RecognizeGeneralDocument"}', "clientId": client_id, "signMethod": "SHA3-256", "signNonce": sign_nonce, "timestamp": timestamp, "signature": signature }) print(resp.json()) ``` ### Method 2: **MCP Protocol Integration** AI Agent collaboration mode, configure once and obtain **32 capabilities**. Supports mainstream AI development tools. **Server URL**: ``` https://scan-mcp.quark.cn/mcp?aiApiKey=your-api-key ``` **Claude Code Integration**: ```bash claude mcp add --transport http quark-scan-business-mcp \ "https://scan-mcp.quark.cn/mcp?aiApiKey=your-api-key" ``` **Cline Integration**: Install Cline extension in VS Code → MCP Server configuration → Fill in Server URL → Select Streamable HTTP protocol. ### Method 3: **Skill Integration** OpenClaw and other Agent Skill integration, the lightest, 3 skill packages covering all capabilities: | Skill Package | Covered Capabilities | Download Address | |----------|---------|---------| | OCR Recognition | General / Handwritten / Table / ID Card / Ticket, etc. | `https://cdn.sm.cn/static/yescan-ocr-universal.zip` | | Document Scanning | Document / ID Card / Contract, etc. scanning enhancement | `https://cdn.sm.cn/static/yescan-scan-universal.zip` | | Office Conversion | Image to Word / Excel / PDF | `https://cdn.sm.cn/static/yescan-transoffice-universal.zip` | ### Method 4: **CLI Command Line Integration** Developer-friendly command-line tool, suitable for script processing and quick testing. ```bash # Installation pip install yescan # Configure API Key yescan config set SCAN_WEBSERVICE_KEY your-api-key # Usage - One-click call by scene name yescan -s general-ocr -p ./document.jpg # General OCR yescan -s idcard-ocr -p ./id_card.jpg # ID Card Recognition yescan -s image-to-word -p ./paper.jpg -o ./out/ # Image to Word yescan -s remove-watermark -p ./watermarked.jpg # Remove Watermark yescan --list-scenes # View all available scenes ``` ### Method 5: **Hermes Agent Integration** Through Hermes Agent skill package mode integration, send pictures + instructions in the dialog box to use. ```bash # Install skill /path/to/yescan-ocr-universal.zip Install this skill # Configure environment variables echo 'SCAN_WEBSERVICE_KEY=your-api-key' >> ~/.hermes/.env # Usage https://img.example.com/doc.jpg Extract text information ``` ### Integration Method Comparison | Characteristics | API | MCP | OpenClaw | CLI | Hermes | |------|-----|-----|----------|-----|--------| | Suitable Scenarios | Business system integration | AI Agent collaboration | Lightweight / Agent | Script / Testing | Conversational | | Development Threshold | Medium | Low | Low | Low | Low | | Capability Coverage | All 36 items | 32 items | All (by package) | All | All (by package) | | Requires Coding | ✅ | ❌ Configuration ready | ❌ | ✅ Command line | ❌ | | Custom Parameters | ✅ Fully controllable | ⚠️ Limited | ⚠️ Limited | ✅ Fully controllable | ⚠️ Limited | --- ## User Scenarios ### 📚 **Education Industry** - **Test Paper Digitalization**: Test paper enhancement + Remove Handwriting → Clean electronic version of test paper - **Question Recognition**: Automatically split and recognize each question from test paper images - **Formula Recognition**: Math formula conversion to LaTeX, automatic entry into question bank - **Homework Correction**: Handwritten text recognition + Remove Handwriting, separate student answers from original questions ### 🏦 **Financial Industry** - **Ticket Recognition**: VAT invoice, train ticket, etc. automatic entry into reimbursement system - **Business License Review**: Automatic extraction of 14 fields, enterprise access automation - **Contract Management**: Contract organization + General text recognition + Image to PDF - **Identity Authentication**: ID card recognition + Face comparison ### 🏛️ **Government Industry** - **Document Digitalization**: ID card, driver's license, social security card, etc. batch recognition - **Document Archiving**: Scanned file + Remove Shadow + Image Quality Enhancement → High-quality electronic archives - **Table Data Entry**: Table recognition → Excel, exempt from manual entry ### 🏥 **Medical Industry** - **Medical Report Digitalization**: Structured extraction of medical report and drug inspection report - **Prescription Recognition**: Handwritten prescription text recognition - **Medical Record Archiving**: Scanning + Remove Background + OCR → Searchable electronic medical record ### 🛒 **E-commerce and Retail** - **Commodity Information Extraction**: Commodity picture text + Barcode recognition - **Product Image Watermark Removal**: Remove competitor watermarks, obtain clean materials - **Picture Translation**: Cross-border e-commerce product image multi-language translation ### 🚗 **Vehicle and Logistics** - **Driver's License / Vehicle Registration**: Vehicle information automatic entry - **Overseas Ticket**: Cross-border logistics document recognition - **Cargo Document Scanning**: Scanned file + Correction enhancement ### 👤 **Personal / HR** - **ID Photo Generation**: 9 styles × 3 color backgrounds, one-click generation for entry / visa / exam - **Resume Scanning**: Paper resume → Electronic editable document - **Note Digitalization**: Handwritten note recognition + Image to Word --- ## Customer Cases ### Case 1: **Education Technology Company - Automated Question Bank Construction** **Pain Point**: Tens of thousands of test papers need manual entry into question bank, extremely inefficient. **Solution**: 1. Test paper enhancement → Remove Handwriting (remove student answers) → General OCR recognition 2. Formula recognition → LaTeX format direct entry 3. Table recognition → Answer card automatic scoring **Effect**: Question bank construction efficiency **10x**, labor cost reduction **80%**. ### Case 2: **Financial Institution - Automated Ticket Reimbursement** **Pain Point**: Monthly processing of tens of thousands of reimbursement invoices, high manual entry error rate. **Solution**: 1. VAT invoice recognition → Full field automatic extraction 2. Train ticket / Commercial invoice recognition → Travel expense automatic summary 3. Business license recognition → Supplier information automatic review **Effect**: Invoice processing time from **3 minutes / piece** to **3 seconds / piece**, accuracy **99%+**. ### Case 3: **Law Firm - Contract Management Digitalization** **Pain Point**: Huge paper contracts are difficult to read, key information retrieval efficiency is low. **Solution**: 1. Scanned file → Remove Shadow → Image Quality Enhancement → High-definition electronic archive 2. General text recognition → Full content extraction → Full text search 3. Image to PDF → Searchable PDF archiving **Effect**: Contract retrieval time from **hours** to **seconds**. --- ## Billing Scheme ### Billing Model By call count, providing multiple package options, prepaid + quota management. ### Free Experience Register and give free experience quota, zero-cost try all capabilities. ### Package Classification | Package | Included Capabilities | Description | |--------|----------------------|--------------| | **OCR Recognition Package** | 19 text recognition capabilities | General text, ID, invoice, formula, etc. | | **File Scanning Package** | 13 scanning enhancement capabilities | Scanning, handwritten removal, watermark removal, enhancement, etc. | | **Format Conversion Package** | 3 format conversion capabilities | Image to Word / Excel / PDF | | **AIGC Package** | 1 AI generation capability | ID photo generation | ### How to Purchase 1. Log in to [Developer Backend](https://scan.quark.cn/business) 2. Go to "Management Center" → "Purchase Packages" 3. Select the required package and specifications 4. Supports Alipay / WeChat Pay --- ## Quick Start ### Step 1: Obtain API Key 1. Visit https://scan.quark.cn/business 2. Click "Developer Backend", register/login with mobile number 3. Obtain **API Key ID** (clientId) and **API Key** (clientSecret) in the backend ### Step 2: Choose Integration Method | Your Needs | Recommended Method | |---------|--------------------| | I have a backend development team | Direct API integration | | I'm using AI Agent (Claude/Cursor/Cline) | MCP integration | | I want to experience it quickly | CLI integration (`pip install yescan`) | | I'm using OpenClaw / Hermes | Corresponding Agent integration | ### Step 3: Make a Call Using CLI as an example, 3 commands can be completed: ```bash pip install yescan yescan config set SCAN_WEBSERVICE_KEY your-api-key yescan -s general-ocr -p ./test.jpg ``` ### Image Format Requirements | Requirement | Specifications | |------|----------------| | Supported formats | PNG / JPG / JPEG / BMP / GIF / TIFF / WebP | | Image size | 15px < length/width < 8192px, aspect ratio < 50 | | File size | ≤ 10MB, recommended 1.5MB | --- ## Frequently Asked Questions ### Q: Can one API Key call all capabilities? A: Yes, all 36 capabilities share the same API Key, and different capabilities are switched through `serviceOption` + `function_option` parameters. ### Q: Does it support PDF input? A: Some capabilities support PDF input (such as general text recognition), and specific capabilities are documented. Image input is the main recommended method. ### Q: What is the QPS limit? A: Different QPS quotas are available based on package types. If high concurrency is required, please contact the business team for customized solutions. ### Q: How is data security ensured? A: SHA3-256 signature authentication, HTTPS encryption in the transmission layer, support for file non-disk options, and compliance with enterprise-level data security regulations. ### Q: How to troubleshoot errors? A: Please refer to the error code table: | Error Code | Meaning | Solution | |--------|------|---------| | `00000` | Success | — | | `A0100` | Unauthorized access | Check if API Key is correct | | `A0202` | Invalid timestamp | Check client time synchronization | | `A0300` | Exceeding QPS limit | Reduce frequency or upgrade package | | `A0301` | Exceeding daily call limit | Wait for the next day to reset or upgrade | | `A0400` | Image parsing failed | Check image format and size | | `B0101` | System timeout | Retry | --- ## Contact Us | Channel | Information | |------|------| | 🌐 Official Website | https://scan.quark.cn/business | | 📞 Cooperation Hotline | 4009030958 | | 📧 Cooperation Email | quark-ai@service.alibaba.com | | 📖 Development Documentation | https://scan.quark.cn/blm/scank-business-docs-703/docs-v2 | * 2025 Alibaba Intelligent Information Business Group · Quark Scan King Team *