/*!
* jbCorrelation()
*
* Input: dirin (directory of input images)
* thresh (typically ~0.8)
* weight (typically ~0.6)
* components (JB_CONN_COMPS, JB_CHARACTERS, JB_WORDS)
* rootname (for output files)
* firstpage (0-based)
* npages (use 0 for all pages in dirin)
* renderflag (1 to render from templates; 0 to skip)
* Return: 0 if OK, 1 on error
*
* Notes:
* (1) The images must be 1 bpp. If they are not, you can convert
* them using convertFilesTo1bpp().
* (2) See prog/jbcorrelation for generating more output (e.g.,
* for debugging)
*/
l_int32
jbCorrelation(const char *dirin,
l_float32 thresh,
l_float32 weight,
l_int32 components,
const char *rootname,
l_int32 firstpage,
l_int32 npages,
l_int32 renderflag) {
char filename[L_BUF_SIZE];
l_int32 nfiles, i, numpages;
JBDATA *data;
JBCLASSER *classer;
PIX *pix;
PIXA *pixa;
SARRAY *safiles;
PROCNAME("jbCorrelation");
if (!dirin)
return ERROR_INT("dirin not defined", procName, 1);
if (!rootname)
return ERROR_INT("rootname not defined", procName, 1);
if (components != JB_CONN_COMPS && components != JB_CHARACTERS &&
components != JB_WORDS)
return ERROR_INT("components invalid", procName, 1);
safiles = getSortedPathnamesInDirectory(dirin, NULL, firstpage, npages);
nfiles = sarrayGetCount(safiles);
/* Classify components */
classer = jbCorrelationInit(components, 0, 0, thresh, weight);
jbAddPages(classer, safiles);
/* Save data */
data = jbDataSave(classer);
jbDataWrite(rootname, data);
/* Optionally, render pages using class templates */
if (renderflag) {
pixa = jbDataRender(data, FALSE);
numpages = pixaGetCount(pixa);
if (numpages != nfiles)
fprintf(stderr, "numpages = %d, nfiles = %d, not equal!\n",
numpages, nfiles);
for (i = 0; i < numpages; i++) {
pix = pixaGetPix(pixa, i, L_CLONE);
snprintf(filename, L_BUF_SIZE, "%s.%05d", rootname, i);
fprintf(stderr, "filename: %s\n", filename);
pixWrite(filename, pix, IFF_PNG);
pixDestroy(&pix);
}
pixaDestroy(&pixa);
}
sarrayDestroy(&safiles);
jbClasserDestroy(&classer);
jbDataDestroy(&data);
return 0;
}
main(int argc,
char **argv)
{
char filename[BUF_SIZE];
char *dirin, *rootname, *fname;
l_int32 reduction, i, firstpage, npages, nfiles;
l_float32 thresh, weight;
JBDATA *data;
JBCLASSER *classer;
NUMA *natl;
SARRAY *safiles;
PIX *pix;
PIXA *pixa, *pixadb;
static char mainName[] = "jbwords";
if (argc != 6 && argc != 8)
exit(ERROR_INT(
" Syntax: jbwords dirin reduction thresh weight rootname [firstpage, npages]",
mainName, 1));
dirin = argv[1];
reduction = atoi(argv[2]);
thresh = atof(argv[3]);
weight = atof(argv[4]);
rootname = argv[5];
if (argc == 6) {
firstpage = 0;
npages = 0;
}
else {
firstpage = atoi(argv[6]);
npages = atoi(argv[7]);
}
classer = jbWordsInTextlines(dirin, reduction, MAX_WORD_WIDTH,
MAX_WORD_HEIGHT, thresh, weight,
&natl, firstpage, npages);
/* Save and write out the result */
data = jbDataSave(classer);
jbDataWrite(rootname, data);
#if RENDER_PAGES
/* Render the pages from the classifier data, and write to file.
* Use debugflag == FALSE to omit outlines of each component. */
pixa = jbDataRender(data, FALSE);
npages = pixaGetCount(pixa);
for (i = 0; i < npages; i++) {
pix = pixaGetPix(pixa, i, L_CLONE);
snprintf(filename, BUF_SIZE, "%s.%05d", rootname, i);
fprintf(stderr, "filename: %s\n", filename);
pixWrite(filename, pix, IFF_PNG);
pixDestroy(&pix);
}
pixaDestroy(&pixa);
#endif /* RENDER_PAGES */
#if RENDER_DEBUG
/* Use debugflag == TRUE to see outlines of each component. */
pixadb = jbDataRender(data, TRUE);
/* Write the debug pages out */
npages = pixaGetCount(pixadb);
for (i = 0; i < npages; i++) {
pix = pixaGetPix(pixadb, i, L_CLONE);
snprintf(filename, BUF_SIZE, "%s.db.%05d", rootname, i);
fprintf(stderr, "filename: %s\n", filename);
pixWrite(filename, pix, IFF_PNG);
pixDestroy(&pix);
}
pixaDestroy(&pixadb);
#endif /* RENDER_DEBUG */
jbClasserDestroy(&classer);
jbDataDestroy(&data);
numaDestroy(&natl);
return 0;
}
int main(int argc,
char **argv)
{
char filename[BUF_SIZE];
char *dirin, *rootname, *fname;
l_int32 i, firstpage, npages, nfiles;
l_float32 thresh, weight;
JBDATA *data;
JBCLASSER *classer;
SARRAY *safiles;
PIX *pix, *pixt;
PIXA *pixa, *pixadb;
static char mainName[] = "jbcorrelation";
if (argc != 5 && argc != 7)
return ERROR_INT(" Syntax: jbcorrelation dirin thresh weight "
"rootname [firstpage, npages]", mainName, 1);
dirin = argv[1];
thresh = atof(argv[2]);
weight = atof(argv[3]);
rootname = argv[4];
if (argc == 5) {
firstpage = 0;
npages = 0;
}
else {
firstpage = atoi(argv[5]);
npages = atoi(argv[6]);
}
#if 0
/*--------------------------------------------------------------*/
jbCorrelation(dirin, thresh, weight, COMPONENTS, rootname,
firstpage, npages, 1);
/*--------------------------------------------------------------*/
#else
/*--------------------------------------------------------------*/
safiles = getSortedPathnamesInDirectory(dirin, NULL, firstpage, npages);
nfiles = sarrayGetCount(safiles);
sarrayWriteStream(stderr, safiles);
/* Classify components on requested pages */
startTimer();
classer = jbCorrelationInit(COMPONENTS, 0, 0, thresh, weight);
jbAddPages(classer, safiles);
fprintf(stderr, "Time to generate classes: %6.3f sec\n", stopTimer());
/* Save and write out the result */
data = jbDataSave(classer);
jbDataWrite(rootname, data);
fprintf(stderr, "Number of classes: %d\n", classer->nclass);
/* Render the pages from the classifier data.
* Use debugflag == FALSE to omit outlines of each component. */
pixa = jbDataRender(data, FALSE);
/* Write the pages out */
npages = pixaGetCount(pixa);
if (npages != nfiles)
fprintf(stderr, "npages = %d, nfiles = %d, not equal!\n",
npages, nfiles);
for (i = 0; i < npages; i++) {
pix = pixaGetPix(pixa, i, L_CLONE);
snprintf(filename, BUF_SIZE, "%s.%05d", rootname, i);
fprintf(stderr, "filename: %s\n", filename);
pixWrite(filename, pix, IFF_PNG);
pixDestroy(&pix);
}
#if DISPLAY_DIFFERENCE
fname = sarrayGetString(safiles, 0, 0);
pixt = pixRead(fname);
pix = pixaGetPix(pixa, 0, L_CLONE);
pixXor(pixt, pixt, pix);
pixWrite("junk_output_diff", pixt, IFF_PNG);
pixDestroy(&pix);
pixDestroy(&pixt);
#endif /* DISPLAY_DIFFERENCE */
#if DEBUG_TEST_DATA_IO
{ JBDATA *newdata;
PIX *newpix;
PIXA *newpixa;
l_int32 same, iofail;
/* Read the data back in and render the pages */
newdata = jbDataRead(rootname);
newpixa = jbDataRender(newdata, FALSE);
iofail = FALSE;
for (i = 0; i < npages; i++) {
pix = pixaGetPix(pixa, i, L_CLONE);
newpix = pixaGetPix(newpixa, i, L_CLONE);
pixEqual(pix, newpix, &same);
if (!same) {
iofail = TRUE;
fprintf(stderr, "pix on page %d are unequal!\n", i);
}
pixDestroy(&pix);
pixDestroy(&newpix);
}
if (iofail)
fprintf(stderr, "read/write for jbdata fails\n");
else
fprintf(stderr, "read/write for jbdata succeeds\n");
jbDataDestroy(&newdata);
pixaDestroy(&newpixa);
}
#endif /* DEBUG_TEST_DATA_IO */
#if RENDER_DEBUG
/* Use debugflag == TRUE to see outlines of each component. */
pixadb = jbDataRender(data, TRUE);
/* Write the debug pages out */
npages = pixaGetCount(pixadb);
for (i = 0; i < npages; i++) {
pix = pixaGetPix(pixadb, i, L_CLONE);
snprintf(filename, BUF_SIZE, "%s.db.%05d", rootname, i);
fprintf(stderr, "filename: %s\n", filename);
pixWrite(filename, pix, IFF_PNG);
pixDestroy(&pix);
}
pixaDestroy(&pixadb);
#endif /* RENDER_DEBUG */
#if DISPLAY_ALL_INSTANCES
/* display all instances, organized by template */
pix = pixaaDisplayByPixa(classer->pixaa,
X_SPACING, Y_SPACING, MAX_OUTPUT_WIDTH);
pixWrite("output_instances", pix, IFF_PNG);
pixDestroy(&pix);
#endif /* DISPLAY_ALL_INSTANCES */
pixaDestroy(&pixa);
sarrayDestroy(&safiles);
jbClasserDestroy(&classer);
jbDataDestroy(&data);
/*--------------------------------------------------------------*/
#endif
return 0;
}
