character said:
Readiris Corporate 12 Asian is available for the Mac; I think it's a leading OCR package.
Looks pretty good, though since they're charging ~10x what we are, we might be able to offer a simpler / cheaper OCR app at a price point Mac users would be happier with - even without all the fancy document management features, a $30 or $40 Chinese OCR app on Mac seems like it could generate a
lot of interest. It's a shame that there's no autofocus in the cameras they build into Macs, though I guess that could be fixed with a high-quality USB webcam (if any still exist...)
character said:
Can I get in on the beta? You would think Apple won the PC/Mac war given how many Macs there are on campus here. If you target PlecOS X at academia, it could do quite well.
Should be possible, yeah - the only closed beta we've ever done is on iPhone and that's purely due to the no-unapproved-apps problem, we're planing to make the Android beta open to all comers and likewise for future any desktop version.
radioman said:
On the topic of OCR, there are packages like that mentioned. Adobe Acrobat is another. The issue is that they are expensive and require processing, importing, etc.
Which is why the quick-and-dirty "app"-like OCR system might make sense.
radioman said:
The other day I took a picture of a crumpled piece of paper on the wall. Poor pix and bad contrast. It was one of those things where there was a announcement of work to be done in the building. Well, I let Pleco have at it, and it got about 70 percent. I was quite impressed. I don't think any of the other Programs could have done better.
It's a pretty spectacular algorithm, especially given its tiny storage / RAM / processor footprint - beats the pants off of a lot of the desktop-based Chinese OCR algorithms we've tried.
radioman said:
My thought is that It is much more valuable to me to have Pleco be able to read pictures of things, snapshots, PDFs or anything that you can extract resolution from the screen, and then OCR on the fly. Rather than trying to use it to read live pix for a few characteers, I know live can be useful, but usually I'm trying to digest more than a few characters at a time.
Well we certainly do plan to keep expanding on the still image capture mode - I think live OCR is a better "hook," and less likely to disappoint people accuracy-wise (still image OCR can sometimes give you a totally mis-recognized document, and that's an easy recipe for a 1-star review), but we licensed the OCR system originally not even knowing if we'd be able to video with it, so still capture has been / should continue to be part of our plans.
radioman said:
You could then have it so that if the document is really clear, the interface would present itself more or less like like the document reader for all intents and purposes. Otherwise you would have some sort of interface like HWR for accuracy that are more challenged.
We're waiting to see how consistent the OCR system's match accuracy statistics are (in the forthcoming update that's adding them) - changing the interface based on accuracy might make sense, but at some point it may be easier to just re-take the photo with better light / straighter angle / etc.
numble said:
So... Mac App Store?
I think the iOS App Store shows that there are definite benefits to "first movers" for software that shows up on these Apple stores.
No way - it may be a necessary evil on smartphones, but with a full keyboard with which to enter one's billing info and an incredibly easy installation process there just isn't enough benefit to justify giving Apple a 30% cut of a very expensive app.
If I thought the market for desktop Chinese dictionary apps was large and lucrative enough that we had to worry about someone getting to it ahead of us, we'd already have a desktop version of Pleco - I still see Pleco Desktop more as an interesting little side product / useful companion to our mobile app than as a crucial part of our business. (particularly with Wenlin 4 rumored to be just around the corner...)