Advanced Development Tools

Useful tools that enhance the developer's ability to interact with ABBYY FineReader Engine and manipulate the recognition process on the core level:

Voting API support
When FineReader Engine is used as one of the participating recognition engines in a third-party application, it supplies recognition alternatives (or hypotheses) with a relevant confidence level for characters, words and intercharacter separation. This information helps developers design an efficient and accurate voting algorithm for applications that require multiple recognition technologies. For example, when recognizing an "O", ABBYY FineReader Engine may return 3 hypotheses: "0" (zero), with 60% confidence; capital "O", with 80% confidence; and capital "C", with 10% confidence. Another example for intercharacter separation: the possible hypotheses for an "m" would be "m", "rn", and "in".

"On-the-fly" tuning of core recognition
ABBYY FineReader provides developers with the access and ability to manipulate the recognition engine during the OCR process on a core level. The FineReader recognition engine generates hypotheses (or recognition alternatives) and allows developers to influence or fine-tune the procedure of setting the confidence level for each hypothesis (or selecting the best hypothesis) using their own specific ranking criteria.

Sample Codes for common conversion tasks
The SDK is supplied with the database of common Engine Usage Samples which help tune FineReader Engine for each particular project in the most appropriate way. This is a set of "ready-to-load" profiles with the optimal speed and accuracy performance balance. The profiles are designed for particular tasks, such as field-level recognition, archiving with imaging and indexing (e.g. searchable PDFs), full-page conversion to RTF and HTML, etc. The database also contains sample images and benchmarks.

<< Back