Enhancing Optical Character Recognition on Images with Mixed Text Using Semantic Segmentation.
Shruti PatilVaradarajan VijayakumarSupriya V. MahadevkarRohan AthawadeLakhan MaheshwariShrushti KumbhareYash GargDeepak DharraoPooja KamatKetan KotechaPublished in: J. Sens. Actuator Networks (2022)
Keyphrases
- optical character recognition
- historical manuscripts
- semantic segmentation
- street scenes
- printed documents
- text extraction
- text regions
- printed text
- image database
- ocr systems
- input image
- text lines
- image analysis
- character recognition
- image understanding
- image features
- multiple images
- object recognition
- natural scenes
- image collections
- test images
- keypoints
- image annotation
- image set
- lighting conditions
- image matching
- historical documents
- segmentation method
- image classification
- image retrieval
- computer vision
- bounding box
- document images