Tesseract on Linux crashes Glassfish

Question

We are using Tess4J/Tesseract to perform OCR on a webapp. On Windows everyting works fine but when deployed on a Linux machine the program crashes, kills the glassfish process and outputs a dump file: hs_err_pidXXXXX.log .

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00007f9fdd5322a0, pid=10412, tid=140324597778176
#
# JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build 1.7.0_75-b13)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C  [libtesseract.so+0x2532a0]  ERRCODE::error(char const*, TessErrorLogCode, char const*, ...) const+0x190
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# If you would like to submit a bug report, please visit:
#   http://bugreport.sun.com/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

---------------  T H R E A D  ---------------

Current thread (0x00007fa00c42d800):  JavaThread "pool-26-thread-1" [_thread_in_native, id=10705, stack(0x00007f9fddbdc000,0x00007f9fddcdd000)]

siginfo:si_signo=SIGSEGV: si_errno=0, si_code=1 (SEGV_MAPERR), si_addr=0x0000000000000000

The tesseract commands works and correctly converts images to text. We have tried the LC_NUMERIC solution but still doesn't work.

Our Tesseract java code is something like this

File file; // ...
boolean hOcr; // ...
Rectangle rec; // ...
OcrResult result;
//Tesseract instance = Tesseract.getInstance();
Tesseract1 instance = new Tesseract1();
try {
    instance.setHocr(hOcr);            
    ImageIO.scanForPlugins();
    String res;
    if (rec == null) {
        res = instance.doOCR(file);
    } else {
        res = instance.doOCR(file, rec);
    }
    result = new OcrResult(res, 0, true);
} catch (TesseractException e) {
    log.error("error tesseract", e);
    // process error
} catch (Error e) {
    log.error("error tesseract", e);
    // process error
}

Our specs

• Tesseract 3.02.02
• Tess4J
• CentoOS 6.4
• Java 1.7
• Glassfish 4.1

Does anyone have any suggestions?

Answer 1

It turned out to be a combination of factors:

• setting datapath to TESSDATA_PREFIX on server JVM settings in Glassfish
• and most importantly, applying patches on Tesseract ( found here , credits to the author) due to a known issue concerning system locale - somehow the bug fixes were not applied in latest versions