OCR, or optical character recognition, is one of life’s disappointments. Like unwrapping a solid, heavy Christmas present and finding eight pairs of socks inside, using OCR with the expectation that what was printed on the paper will actually appear in perfect form on your PC screen usually results in amazement, followed swiftly by annoyance.
This is especially true if you use one of the ‘lite’ packages that’s supplied with the majority of scanners. There are better packages on the market, though, and one of them is Presto! OCR Pro 4.0. This is OCR ‘with human-eye precision’, apparently. It doesn’t say whether that’s the 20:20 vision of someone in their early 20s or the rheumy old squint of a pensioner with bottle-bottom specs, but judging from our tests, it’s nearer the former.
Initially we played with the sample TIF files supplied with the application and – as you’d expect – these produced some impressive results, with good tabulation, columns and images. The TIF-handling feature means that you can use the package to convert electronic faxes into conventional documents. And, if you’re feeling clever, you can do it with source documents from dozens of different languages, even if there are multiple languages on a single page.
Things get better. Scanning a relatively complex document resulted in excellently captured fonts, tables and diagrams. The company claims that price lists, inventory lists, schedules and other tabulated documents are accurately recognised, and that all seems to be true. Direct export to Word and Excel (with WordPro and WordPerfect also supported) results in well-formatted tables that need little correction. Inevitably, there’s also an HTML export option, and this one works pretty well – ideal for anyone looking to republish paper documents on the Web or an Intranet.
Also inevitably, errors do creep in, but the package allows you to use its built-in dictionaries to run through the text, highlighting and correcting any errors. Supposedly, like a voice-recognition package, the program will learn from these mistakes with unrecognised fonts and do a better recognition job next time around.
Contact: 01752 895100