How to increase the processing speed for a big batch of images
To speed up the big batch processing you can definitely parallel the processing in the following cases:
- to process one-page documents;
- to perform processing using the document definitions which are not created from the AFL (ABBYY Flexible Layout) file in the batch, as each page in this situation can be recognized independently;
- to process the documents with the known borders.
More Complicated Case
In case there are a lot of pages from the multi-page documents and at least one document definition with the flexible section, then to parallel such batches is more complicated, as the process of the document definition applying for the next page will be depended on the previous one.
- If it is possible, break the batch up to the parts, process them in parallel and concatenate them into the one batch.
- If the customer uses a lot of documents definitions, then use the classification tree in the project.
- Also, you can make flexible one-page document definition which will search specific objects on the whole pages and depending on it will identify the type of the document for the single image. Such processing can be parallel. After this, the documents can be grouped according to the results of the first step and then recognize the obtained groups in parallel. This way works like a classification tree and can be considered as its' alternative.
General recommendations about increasing processing speed in FlexiCapture Engine can be found in the corresponding Help article: "Help" → "Guided Tour" → "Best Practices" → "Increasing Processing Speed".
Please sign in to leave a comment.