Clean scanned PDFs by classifying page types and optimizing cleaning parameters.
npx skills add https://github.com/imvladikon/dot-claude --skill clean-scan
Please help me install this Agent Skill. Skill: clean-scan Source: https://github.com/imvladikon/dot-claude/tree/main/skills/clean-scan Command: npx skills add https://github.com/imvladikon/dot-claude --skill clean-scan
Очистка сканов PDF от серого и желтоватого фона с адаптивной подгонкой параметров под конкретный скан.
Активируйте виртуальное окружение и запустите adaptive_cleaner.py с входным PDF и выходным файлом.
High-intent search queries and answers about installing and using this skill.
Adaptive PDF scan cleaning removes gray or yellow backgrounds by classifying scan types like YELLOW_AGED or UNEVEN_LIGHTING and optimizing parameters such as percentile thresholds and whitening levels per page. This automated process ensures clean, high-contrast output for OCR preprocessing.
Automating scan cleaning for a multi-page PDF involves analyzing several pages to agree on consistent cleaning parameters across the entire document. Using a command-line interface with an --auto flag, the system applies per-page feature analysis to select the optimal kernel size and sharpness settings globally.
OpenCV supports adaptive background removal for unevenly lit PDF scans by applying per-page feature analysis and scan-type classification. The process adjusts parameters like percentile_low and percentile_high, and can apply CLAHE to effectively whiten uneven backgrounds and enhance readability.
You can skip specific pages during automated PDF scan cleaning by using the --skip-pages CLI parameter. This allows the adaptive cleaner to bypass unwanted pages, applying the optimized background removal and whitening threshold only to the targeted sections of the document.
To test scan cleaning parameters on a single page before processing a full PDF, use the --test-page CLI parameter. This runs the adaptive parameter optimization and feature analysis on one specified page, allowing you to verify kernel size and sharpness adjustments before full execution.
A PDF scan may still show gray backgrounds after automated cleaning if the scan-type classification mismatched the page features or if the whitening_threshold was insufficient. Adjusting the --dpi setting or enabling --clahe can improve contrast and correct uneven lighting artifacts.
FnSK4R17sExecute Python scripts in a temporary virtual environment with dependencies.
89jobrienExtract text and tables from native and scanned PDFs using pdfplumber.
jiyangnanConvert PDF documents to editable Word files with Chinese OCR.
z4nr34lAnalyze website SEO health and output prioritized recommendations.
