Check autorotate using leptonica correlation

ocrmypdf/leptonica.py

@@ -193,6 +193,14 @@ def deskew(infile, outfile, dpi):
         raise LeptonicaIOError("Failed to open destination file: %s" % outfile)
 
 
+def correlation_binary(pix1, pix2):
+    correlation = ffi.new('float *', 0.0)
+    result = lept.pixCorrelationBinary(pix1, pix2, correlation)
+    if result != 0:
+        raise LeptonicaError("Correlation failed")
+    return correlation[0]
+
+
 if __name__ == '__main__':
     parser = argparse.ArgumentParser(
         description="Python wrapper to access Leptonica")

ocrmypdf/lib/compile_leptonica.py

@@ -52,6 +52,7 @@ l_int32 pixWriteImpliedFormat ( const char *filename, PIX *pix, l_int32 quality,
 void pixDestroy ( PIX **ppix );
 PIX * pixDeskew ( PIX *pixs, l_int32 redsearch );
 char * getLeptonicaVersion (  );
+l_int32 pixCorrelationBinary(PIX *pix1, PIX *pix2, l_float32 *pval);
 """)
 
 if __name__ == '__main__':

tests/test_main.py

@@ -11,6 +11,7 @@ import pytest
 from ocrmypdf.pageinfo import pdf_get_all_pageinfo
 import PyPDF2 as pypdf
 from ocrmypdf import ExitCode
+from ocrmypdf import leptonica
 
 
 if sys.version_info.major < 3:
@@ -259,6 +260,11 @@ def test_autorotate(spoof_tesseract_cache, renderer):
             _make_input('cardinal.pdf'), _make_output('reference.png'),
             xres=100, yres=100, raster_device='pngmono', log=gslog, pageno=1)
 
+    pix_ref = leptonica.pixRead(_make_output('reference.png'))
+    for n in range(1, 4+1):
+        pix_other = leptonica.pixRead(_make_output('cardinal-%i.png' % n))
+        print(leptonica.correlation_binary(pix_ref, pix_other))
+
 
 @pytest.mark.parametrize('renderer', [
     'hocr',