The aim of this competition is to find the best system able to read complete words in camera captured scenes.

Example:

We expect a robust reading algorithm to find and recognize the following words:

That is, it should have identified five words: "Department", "of", "Computer", "Science", "1", and also specified a bounding rectangle (in image pixel coordinates).

Note that the Text Locating, Robust Character Recognition and Robust Word Recognition all tackle sub-parts of this problem.  The robust reading competition aims to identify the system that best does the complete job.

Recognizer Interface:  details to follow

Evaluation Interface:   there are two ways your recognizer can interact with our evaluation software: file-based, and API-based.

File-based interaction:   details to follow.

API-based interaction:   details to follow.

Community:   visit the discussion forum to share your views with the document analysis and recognition community.  Have your say on dataset formats, test image characteristics, software interfaces and evaluation procedures.

Sample and Trial Datasets now available