OCR0018: Load transcription work for HTR team (1)

OpenPecha / image-to-text

MIT License

0 stars 0 forks source link

OCR0018: Load transcription work for HTR team (1) #33

Open ta4tsering opened 5 months ago

ta4tsering commented 5 months ago

Description: HTR team is a Team of annotators from nepal who are working on transcribing pechas and their work is finished so need to provide them with new images to trancribe.

Completion Criteria: CSV uploaded to the Database and informing annotators

Reference:

bdrc image object key s3://image-processing.bdrc.io/scam_cropped/PT/Manual2-20240411/

Implementation plan:

Subtask:

[x] Download the images from bdrc s3
[x] upload the images to monlam s3
[x] create csv and import to database

ta4tsering commented 5 months ago

worked on NOFO for another project for three days so didn't finish on time with the estimation as per estimated in ahead

ta4tsering commented 5 months ago

Downloaded and cropped the line images for the Correction-3 and created csv for it. Correction batch 3 ready for the HTR team. Refactored the scripts to make it more easy to use for the next batch. https://github.com/OpenPecha/image-to-text/tree/main/src/HTR

ta4tsering commented 3 months ago

Update the script to change the line name from using pagename_coordinates.jpg to the line image provided in the xml file pagename_linename.jpg and creating the next batch correction-8 which has 18000 pages.