{"categories":[{"label":"Artificial Intelligence","url":"https://skillfed.io/packages/category/scientific-engineering-artificial-intelligence/5"}],"enrichment":{"capability":"Recognizes text and detects objects in captcha images using offline neural network models, supporting digits, letters, Chinese characters, and slider captchas.","skillfed_tags":["captcha-solving","offline-ml","image-recognition"],"use_cases":["Automate login flows that require solving simple alphanumeric captchas in web scraping or testing","Batch-process captcha images offline without hitting rate limits or costs of external OCR services","Detect and locate objects or text regions in images for accessibility or content moderation workflows","Recognize Chinese character captchas or mixed-language verification codes in regional applications","Match slider puzzle gaps by comparing edge patterns between template and target images"],"what_it_does":"ddddocr is an offline, local captcha recognition library that uses pre-trained neural network models to identify text and objects in images. It handles common alphanumeric captchas, Chinese characters, slider puzzles, and special characters without requiring external API calls. The library wraps ONNX Runtime models and depends on numpy, Pillow, and OpenCV for image processing.\n\nYou initialize it once, then call `classification()` on image bytes to get recognized text. It supports multiple modes\u2014standard OCR, object detection, slider matching\u2014and allows GPU acceleration if you have CUDA available. The package also provides probability distributions per character and lets you constrain recognition to specific character sets, useful when you know the captcha format in advance.","worth_installing":"Yes, if you need offline captcha recognition and accept the trade-off that accuracy depends on model fit to your specific captcha style. The permissive MIT license, active maintenance, low install friction, and no known vulnerabilities make it a reasonable choice for development and production. Start with the default model; if accuracy is poor, try the beta model or train a custom one."},"id":"ddddocr","links":{"html":"https://skillfed.io/packages/ddddocr","md":"https://skillfed.io/packages/ddddocr.md","pypi":"https://pypi.org/project/ddddocr/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-03-11","license_spdx":null,"license_treatment":"permissive","name":"ddddocr","python_support":"supports_current","summary":"\u5e26\u5e26\u5f1f\u5f1fOCR"},"popularity":{"monthly_downloads":416565,"position":6819,"tier":"top_15000"},"security":{"n_vulnerabilities":0},"version":"1.6.1"}