optical character recognisation

docs/OCR.txt

@@ -0,0 +1,96 @@
+
+Overview
+========
+OCR (Optical Character Recognisation ) is an technique used to 
+extract text from images. In the World of Subtile, subtitle stored 
+in bitmap format are common and even neccassary. for converting subtile 
+in bitmap format to subtilte in text format ocr is used.
+
+Dependecy
+=========
+Tesseract (OCR library by google)
+Leptonica (image processing library)
+
+How to compile ccextractor  on linux with OCR
+=============================================
+
+Download and Install Leptonnica.
+-------------------------------
+This package is available, you need liblept-devel library.
+
+If Leptonica isn't available for your distribution, or you want to use a newer version
+ than they offer, you can compile your own.
+
+you can download lib leptonica from  http://www.leptonica.com/download.html
+
+Download and Install Tesseract.
+-------------------------------
+Tesseract is available directly from many Linux distributions. The package is generally
+ called 'tesseract' or 'tesseract-ocr' - search your distribution's repositories to
+ find it. Packages are also generally available for language training data (search the
+ repositories,) but if not you will need to download the appropriate training data,
+ unpack it, and copy the .traineddata file into the 'tessdata' directory, probably
+ /usr/share/tesseract-ocr/tessdata or /usr/share/tessdata.
+
+If Tesseract isn't available for your distribution, or you want to use a newer version
+ than they offer, you can compile your own.
+
+If you compile Tesseract then following command in its source code are enough
+./autogen.sh
+./configure
+make
+sudo make install
+sudo ldconfig
+
+ Note: 
+1) CCExtractor is tested with Tesseract 3.02.02 version. 
+
+you can download tesseract from https://drive.google.com/folderview?id=0B7l10Bj_LprhQnpSRkpGMGV2eE0&usp=sharing
+
+
+
+Compile CCextractor passing flags like following
+-------------------------------------------------
+make ENABLE_OCR=yes
+
+
+How to compile ccextractor  on Windows with OCR
+=============================================
+
+Download prebuild library of leptonica from following link
+http://www.leptonica.com/source/leptonica-1.68-win32-lib-include-dirs.zip
+
+Download prebuild library of tesseract from following tesseract official link
+https://code.google.com/p/tesseract-ocr/downloads/detail?name=tesseract-3.02.02-win32-lib-include-dirs.zip
+
+put the path of libs/include of leptonica and tesseract in library paths.
+step 1) In visual studio 2013 right click <Project> and select property.
+step 2) Select Configuration properties in left panel(column) of property.
+step 3) Select VC++ Directory.
+step 4) In the right pane, in the right-hand column of the VC++ Directory property,
+	open the drop-down menu and choose Edit.
+Step 5) Add path of Directory where you have kept uncompressed library of leptonica
+	and tesseract.
+
+
+Set preprocessor flag ENABLE_OCR=1
+Step 1)In visual studio 2013 right click <Project> and select property.
+Step 2)In the left panel, select Configuration Properties, C/C++, Preprocessor.
+Step 3)In the right panel, in the right-hand column of the Preprocessor Definitions property, open the drop-down menu and choose Edit.
+Step 4)In the Preprocessor Definitions dialog box, add ENABLE_OCR=1. Choose OK to save your changes.
+
+Add library in linker
+step 1)Open property of project
+Step 2)Select Configuration properties
+Step 3)Select Linker in left panel(column)
+Step 4)Select Input
+Step 5)Select Additional dependencies in right panel
+Step 6)Add libtesseract302.lib in new line
+Step 7)Add liblept168.lib in new line
+
+Download language data from following link
+https://code.google.com/p/tesseract-ocr/downloads/list
+after downloading the tesseract-ocr-3.02.eng.tar extract the tar file and put
+tessdata folder where you have kept ccextractor executable
+
+Copy the tesseract and leptonica dll in the folder of executable or in system32.

linux/Makefile

@@ -35,6 +35,11 @@ INSTLALL = cp -f -p
 INSTLALL_PROGRAM = $(INSTLALL)
 DESTDIR = /usr/bin
 
+ifeq ($(ENABLE_OCR),yes)
+CFLAGS+=-I/usr/local/include/tesseract -I/usr/local/include/leptonica
+CFLAGS+=-DENABLE_OCR
+LDFLAGS+= $(shell pkg-config --libs tesseract)
+endif
 .PHONY: all
 all: objs_dir $(TARGET) 
 

src/dvb_subtitle_decoder.c

@@ -33,6 +33,7 @@
 
 #include "dvb_subtitle_decoder.h"
 #include "spupng_encoder.h"
+#include "ocr.h"
 #define DEBUG
 
 #ifdef DEBUG
@@ -255,9 +256,17 @@ int mapclut_paletee(png_color *palette, png_byte *alpha, uint32_t *clut,
 	}
 	return 0;
 }
-
+/*
+ * @param alpha out
+ * @param intensity in 
+ * @param palette out should be already initialized 
+ * @param bitmap in 
+ * @param size in size of bitmap
+ * @param max_color in
+ * @param nb_color in
+ */
 int quantize_map(png_byte *alpha, uint8_t *intensity, png_color *palette,
-		uint8_t *bitmap, int h, int w, int max_color, int nb_color)
+		uint8_t *bitmap, int size, int max_color, int nb_color)
 {
 	/*
 	 * occurrence of color in image
@@ -301,12 +310,9 @@ int quantize_map(png_byte *alpha, uint8_t *intensity, png_color *palette,
 	memset(mcit, 0, nb_color * sizeof(uint32_t));
 
 	/* calculate histogram of image */
-	for (int i = 0; i < h; i++)
+	for (int i = 0; i < size; i++)
 	{
-		for (int j = 0; j < w; j++)
-		{
-			histogram[bitmap[i * w + (j)]]++;
-		}
+		histogram[bitmap[i]]++;
 	}
 	sort_intensity_wise((uint8_t*) alpha, (uint8_t*) intensity, iot, nb_color);
 
@@ -365,52 +371,65 @@ int quantize_map(png_byte *alpha, uint8_t *intensity, png_color *palette,
 	freep(&iot);
 	return ret;
 }
-static int save_spupng(const char *filename, uint8_t *bitmap, int w, int h,
-		uint32_t *clut, uint8_t *luit,uint8_t depth)
-{
-	FILE *f = NULL;
-	png_structp png_ptr = NULL;
-	png_infop info_ptr = NULL;
-	png_bytep* row_pointer = NULL;
-	int i, j, ret;
-	int k = 0;
 
-	png_color *palette = NULL;
-	png_byte *alpha = NULL;
+
+static int pre_process_bitmap(png_color **palette, png_byte **alpha, int size,
+		uint32_t *clut, uint8_t *luit, uint8_t *bitmap, uint8_t depth)
+{
+	/*local pointer to palette */
+	png_color *lpalette = NULL;
+	/* local pointer to alpha */
+	png_byte *lalpha = NULL;
 	int nb_color = (1<< depth);
+	int ret = 0;
 
-	if(!h)
-		h = 1;
-	if(!w)
-		w = 1;
 
-	palette = (png_color*) malloc(nb_color * sizeof(png_color));
-	if(!palette)
+	lpalette = (png_color*) malloc(nb_color * sizeof(png_color));
+	if(!lpalette)
 	{
 		ret = -1;
 		goto end;
 	}
-	alpha = (png_byte*) malloc(nb_color * sizeof(png_byte));
-	if(!alpha)
+	lalpha = (png_byte*) malloc(nb_color * sizeof(png_byte));
+	if(!lalpha)
 	{
 		ret = -1;
 		goto end;
 	}
-
-
 	if(clut)
-		mapclut_paletee(palette, alpha, clut, nb_color);
+		mapclut_paletee(lpalette, lalpha, clut, nb_color);
 	else
 	{
 		/* initialize colors with white */
-		memset(palette,0xff,sizeof(nb_color * sizeof(*palette)));
+		memset(palette,0xff,sizeof(nb_color * sizeof(*lpalette)));
 
 		/* initialize transparency as complete transparent */
-		memset(alpha,0,sizeof(nb_color * sizeof(*alpha)));
+		memset(lalpha,0,sizeof(nb_color * sizeof(*lalpha)));
 	}
 
 	if(bitmap)
-		quantize_map(alpha, luit, palette, bitmap, h, w, 3, nb_color);
+	{
+		quantize_map(lalpha, luit, lpalette, bitmap, size, 3, nb_color);
+	}
+	*palette = lpalette;
+	*alpha = lalpha;
+end:
+	return ret;
+}
+static int save_spupng(const char *filename, uint8_t *bitmap, int w, int h,
+		png_color *palette, png_byte *alpha, int nb_color)
+{
+	FILE *f = NULL;
+	png_structp png_ptr = NULL;
+	png_infop info_ptr = NULL;
+	png_bytep* row_pointer = NULL;
+	int i, j, ret = 0;
+	int k = 0;
+	if(!h)
+		h = 1;
+	if(!w)
+		w = 1;
+
 
 	f = fopen(filename, "wb");
 	if (!f)
@@ -791,6 +810,11 @@ static void save_display_set(DVBSubContext *ctx)
 
 	if (x_pos >= 0)
 	{
+		png_color *palette = NULL;
+		png_byte *alpha = NULL;
+#ifdef ENABLE_OCR
+		char*str = NULL;
+#endif
 
 		filename = get_spupng_filename(sp);
 		inc_spupng_fileindex(sp);
@@ -821,17 +845,28 @@ static void save_display_set(DVBSubContext *ctx)
 			}
 
 		}
-		save_spupng(filename, pbuf, width, height,
-				clut->clut16, clut->ilut16,region->depth);
+		pre_process_bitmap(&palette,&alpha,width*height,clut->clut16, clut->ilut16,pbuf,region->depth);
+#ifdef ENABLE_OCR
+		str = ocr_bitmap(palette,alpha,pbuf,width,height);
+		if(str)
+		{
+			write_spucomment(sp,str);
+		}
+#endif
+		save_spupng(filename, pbuf, width, height, palette, alpha,(1 << region->depth));
 
 		free(pbuf);
+		freep(&palette);
+		freep(&alpha);
 	}
 	else if(!ctx->prev_start)
 	{
+		png_color palette = {0,0,0};
+		png_byte alpha = 0;
 		filename = get_spupng_filename(sp);
 		inc_spupng_fileindex(sp);
 		/* save dummy frame */
-		save_spupng(filename,NULL,1,1,NULL,NULL,0);
+		save_spupng(filename,NULL,1,1,&palette,&alpha,1);
 
 	}
 

src/dvb_subtitle_decoder.h

@@ -86,6 +86,7 @@ int parse_dvb_description(struct dvb_config* cfg, unsigned char*data,
  * 
  */
 void dvbsub_set_write(void *dvb_ctx, struct ccx_s_write *out);
+
 #ifdef __cplusplus
 }
 #endif

src/ocr.c

@@ -0,0 +1,57 @@
+#include "png.h"
+#ifdef ENABLE_OCR
+#include "platform.h"
+#include "capi.h"
+#include "allheaders.h"
+
+char* ocr_bitmap(png_color *palette,png_byte *alpha, unsigned char* indata,int w, int h)
+{
+	TessBaseAPI* api;
+	PIX	*pix;
+	char*text_out= NULL;
+	int i,j,index,ret;
+	unsigned int wpl;
+	unsigned int *data,*ppixel;
+	api = TessBaseAPICreate();
+
+	pix = pixCreate(w, h, 32);
+	if(pix == NULL)
+	{
+		return NULL;
+	}
+	wpl = pixGetWpl(pix);
+	data = pixGetData(pix);
+#if LEPTONICA_VERSION > 69
+	pixSetSpp(pix, 4);
+#endif
+	for (i = 0; i < h; i++)
+	{
+		ppixel = data + i * wpl; 
+		for (j = 0; j < w; j++)
+		{
+			index = indata[i * w + (j)];
+			composeRGBPixel(palette[index].red, palette[index].green,palette[index].blue, ppixel);
+			SET_DATA_BYTE(ppixel, L_ALPHA_CHANNEL,alpha[index]);
+			ppixel++;
+		}    
+	}    
+
+	ret = TessBaseAPIInit3(api,"", "eng");
+	if(ret < 0)
+	{
+		return NULL;
+	}
+
+	//text_out = TessBaseAPIProcessPages(api, "/home/anshul/test_videos/dvbsubtest.d/sub0018.png", 0, 0);
+	text_out = TessBaseAPIProcessPage(api, pix, 0, NULL, NULL, 1000);
+	if(!text_out)
+		printf("\nsomething messy\n");
+	return text_out;
+}
+#else
+char* ocr_bitmap(png_color *palette,png_byte *alpha, unsigned char* indata,unsigned char d,int w, int h)
+{
+	mprint("ocr not supported without tesseract\n");
+	return NULL;
+}
+#endif

src/ocr.h

@@ -0,0 +1,6 @@
+#ifndef OCR_H
+#define OCR_H
+#include <png.h>
+char* ocr_bitmap(png_color *palette,png_byte *alpha, unsigned char* indata,int w, int h);
+
+#endif

windows/ccextractor.sln

@@ -1,18 +1,26 @@
 
-Microsoft Visual Studio Solution File, Format Version 10.00
-# Visual C++ Express 2008
-Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "ccextractor", "ccextractor.vcproj", "{0F0063C4-BCBC-4379-A6D5-84A5669C940A}"
+Microsoft Visual Studio Solution File, Format Version 12.00
+# Visual Studio Express 2013 for Windows Desktop
+VisualStudioVersion = 12.0.21005.1
+MinimumVisualStudioVersion = 10.0.40219.1
+Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "ccextractor", "ccextractor.vcxproj", "{0F0063C4-BCBC-4379-A6D5-84A5669C940A}"
 EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Win32 = Debug|Win32
+		Debug|x64 = Debug|x64
 		Release|Win32 = Release|Win32
+		Release|x64 = Release|x64
 	EndGlobalSection
 	GlobalSection(ProjectConfigurationPlatforms) = postSolution
 		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Debug|Win32.ActiveCfg = Debug|Win32
 		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Debug|Win32.Build.0 = Debug|Win32
+		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Debug|x64.ActiveCfg = Debug|x64
+		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Debug|x64.Build.0 = Debug|x64
 		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Release|Win32.ActiveCfg = Release|Win32
 		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Release|Win32.Build.0 = Release|Win32
+		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Release|x64.ActiveCfg = Release|x64
+		{0F0063C4-BCBC-4379-A6D5-84A5669C940A}.Release|x64.Build.0 = Release|x64
 	EndGlobalSection
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE