Chineseocr

Chinese OCR Tools - Provides optical character recognition specifically for Chinese scripts using deep learning models.

End-to-End Pipelines - Provides an end-to-end pipeline that detects text regions and recognizes Chinese and English characters.

Text Detectors in Images - Locating and bounding text regions in images containing Chinese and English characters using deep learning models like YOLO3 and darknet.

CRNN Recognizers - Provides a sequence-based recognizer that reads variable-length Chinese and English text using CRNN models.

Dense OCR Models - Ships dense OCR models for high-accuracy text extraction from document images.

YOLO-Based Detectors - Uses YOLOv3 object detection architecture to locate text regions in images with bounding boxes.

Deep Learning Detectors - Locates text regions in images using darknet, OpenCV DNN, or Keras deep learning models.

CRNN-Based Recognizers - Provides a CRNN-based recognizer for variable-length Chinese and English text sequences.

Variable-Length Text Recognizers - Recognize variable-length Chinese and English text from detected regions using CRNN or dense OCR models.

Cross-Framework Model Conversion - Converts trained models between darknet, Keras, and PyTorch formats for cross-framework deployment.

Model Format Converters - Converts trained deep learning models between darknet, Keras, and PyTorch formats for cross-framework deployment.

Model Format Converters - Converts trained text detection and recognition models between darknet, Keras, and PyTorch formats.

Text Orientation Detection - Detects text orientation at 0, 90, 180, or 270 degrees using deep learning models before recognition.

Document - Pulls specific fields from identity cards and train tickets using OCR and layout analysis.

Document Field Extractors - Extracts structured fields from identity cards and train tickets using trained models.

chineseocrchineseocr

Features

Star history