Quan Nguyen

It was recently discovered that the `jsane` library, the java wrapper to the `SANE` functions, only works with Linux 32-bit. Running it on Linux 64-bit will throw UnsatisfiedLinkError exceptions. It...

We can recompile the library on Linux Ubuntu 64-bit, but it looks like some change to the source code is necessary as UnsatisfiedLinkError exceptions still happen. Need JNI expert help!

Are you sure this is a valid path `JNA_LIBRARY_PATH+File.pathSeparator+"/opt/homebrew/lib/"` as it would resolve to `"jna.library.path:/opt/homebrew/lib/"` ? Tess4J relies on JNA to find and load native libraries, so we feel this...

The exception appears to have happened deep down in Tesseract native code. Is your application running in single thread or multi-thread? Have you tried using `Tesseract1` class?

You may want to dispose of and start a new instance after so many OCR operations because of memory leaks in native code.

You'd need to preprocess such images. We applied the monochrome filter in [VietOCR](https://github.com/nguyenq/VietOCR3), which uses Tess4J, and were able to get the OCR text. 

Since Tesseract probably has this preprocessing step when reading an image, you'll need to do the same in your Java code as Tess4J wrapper does not include any image preprocessing;...

https://github.com/tesseract-ocr/tessdoc/blob/main/ImproveQuality.md#transparency--alpha-channel

@waljohn Please submit a PR.

We may need to debug and trace through the native code to determine what preprocessing is performed for this kind of image.