OCRKit (US$59.99) has a useful goal: converting photos into searchable PDF files (also RTF, HTML and text). You drag in photos, it produces documents. It couldn't be simpler to use.
An app like this lives and dies by virtue of of its precision, however. Unfortunately, OCRKit did not shine in comparison tests. I created a number of samples and ran them through OCRKit as well as a free online converter tool (newocr.com).
OCRKit consistently produced more errors in its output than the online tool, although the same kinds of errors challenged both approaches ("bLock" vs. "b1ock" vs. "block," for example, or "redcolor" vs. "red color," or "*" vs. "+").
I would have expected the paid tool to outshine the free one but, surprisingly, it did not. This flavored my overall impression about the software.
I did like the app's output. OCRKit combined the scanned image visuals with the interpreted text. It made it easy to search for text and find it within the context of the original document. The product's output is something many users will find of value.
Here is this post, after screenshotting it and passing it through OCRKit. It's been converted to an OCR PDF, enabling me to create selections from the scanned image. The selection you see is not in the original screen shot. It's me selecting the text in Preview from the output PDF.
And here is the text I selected, after copying and pasting that selection to TextEdit:
I created a number of test samples and ran them through OCRKit as well as a free online converter tool (newocr.corn).