Render a page in a correct DPI

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   Hello,

I open and render each page in a byte[] to OCRize them.
But I don't really understand what I need to set as rendered DPI.

            byte[] bytes = Engine.RenderPageToString(RenderedPdfDpi, pageIdx + 1, 0);

If I set 300 for example, and my pdf was in 96, I have a *3 zoom, which is very time consuming for my OCR. At the contrary, if I set 150 and my pdf was in 300, I have bad recognition results.

My PDF are simple multi pages images (like a TIF), so I thought about selecting the first image of each page, and reproduce the same DPI. But this do not work:
Engine.SelectPage(pageIdx + 1);
Engine.SelectImage(Engine.GetImageID(1));

Can somebody help me ?

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
moi meme Members Profile Find Members Posts Beginner Joined: 22 Nov 10 Status: Offline Points: 16	Post Options Post Reply Quote moi meme Report Post Thanks(0) Quote Reply Topic: Render a page in a correct DPI Posted: 09 Mar 11 at 2:20PM
	Hello, I open and render each page in a byte[] to OCRize them. But I don't really understand what I need to set as rendered DPI. byte[] bytes = Engine.RenderPageToString(RenderedPdfDpi, pageIdx + 1, 0); If I set 300 for example, and my pdf was in 96, I have a *3 zoom, which is very time consuming for my OCR. At the contrary, if I set 150 and my pdf was in 300, I have bad recognition results. My PDF are simple multi pages images (like a TIF), so I thought about selecting the first image of each page, and reproduce the same DPI. But this do not work: Engine.SelectPage(pageIdx + 1); Engine.SelectImage(Engine.GetImageID(1)); Can somebody help me ?

moi meme Members Profile Find Members Posts Beginner Joined: 22 Nov 10 Status: Offline Points: 16	Post Options Post Reply Quote moi meme Report Post Thanks(0) Quote Reply Posted: 09 Mar 11 at 2:24PM
	Note: To create the document, I simply load a PDF file with: Engine.LoadFromFile(filename);

moi meme Members Profile Find Members Posts Beginner Joined: 22 Nov 10 Status: Offline Points: 16	Post Options Post Reply Quote moi meme Report Post Thanks(0) Quote Reply Posted: 09 Mar 11 at 2:31PM
	Hum... Now, the image selection works (I must have confused myself). But I'm still unable to get the resolution of the image: ImageHorizontalResolution returns 0. code: Engine.LoadFromFile(filename); Engine.SelectPage(pageIdx + 1); int imgID = Engine.GetImageID(pageIdx + 1) Engine.SelectImage(imgID); int resH = Engine.ImageHorizontalResolution();

moi meme Members Profile Find Members Posts Beginner Joined: 22 Nov 10 Status: Offline Points: 16	Post Options Post Reply Quote moi meme Report Post Thanks(0) Quote Reply Posted: 09 Mar 11 at 3:01PM
	Ok I found :> I only needed to set the right resolution to the rendered bitmap: RenderedPdfDpi = 300; byte[] bytes = Engine.RenderPageToString(RenderedPdfDpi, pageIdx + 1, 0); Bitmap bitmap = new Bitmap(new MemoryStream(bytes)); bitmap.SetResolution(RenderedPdfDpi, RenderedPdfDpi);