Pix and other TessBaseAPI data structures, such as STRING, are not available to Java client code as Java objects. I am only able to use the types that the C wrapper exposes. Moreover, keeping external dependencies to a minimum is also a desirable design objective.
I did put in a fix that has made the image data more consumable to the libtesseract302.dll and, as a result, no longer experience the exceptions I used to. I haven't a chance to test to determine whether the outputs from Pix type vs. raw image type are of similar quality. But if Pix type actually received "preferential" treatment and produced superior recognition quality, IMHO, the raw image should be converted to Pix inside Tesseract, e.g., in SetImage(imagedata) function which subsequently calls SetImage(pix), before processing for recognition. On Monday, April 23, 2012 10:18:20 PM UTC-5, TP wrote: > > On Mon, Apr 23, 2012 at 7:07 PM, Quan Nguyen wrote: > > all of the provided image processing functions are geared for Pix type, > not > > raw image. > > Why not just create a Pix from the raw image data? Leptonica has > pixCreateHeader(), pixSetResolution(), pixSetWpl(), pixSetData(), etc > [1] and various helper functions for byteswapping [2] > > [1] > http://tpgit.github.com/Leptonica/pix1_8c.html#a9a6808a8f168fb1e31cea82bc6ad5d19 > > [2] http://tpgit.github.com/Leptonica/pix2_8c_source.html#l00079 > > -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/tesseract-ocr?hl=en
