How do I overlay text onto an image, regardless of image rotation.

[Ryan]

Q:
I have a document with multiple images, some of which appear to be rotated in various ways (such as upside down).

I want to overlay text onto the image, but I want the text to always be right side up.

Is there a way to do this?

A:
Yes, this is easy to do, by simply finding out the bounding box of the images, and adding text overtop. This works as long as there is no skewing, and if the text is centered on the bounding box, will also work with any rotation angle.

Here is python code, based on ElementBuilder and ImageExtract samples, that creates various images in different rotations, and then overlays text on them.

def ImageFinder(reader, rects):
    element = reader.Next()
    while element != None:
        if (element.GetType() == Element.e_image or
            element.GetType() == Element.e_inline_image):

            rects.append(element.GetBBox())
        elif element.GetType() == Element.e_form:
            reader.FormBegin()
            ImageFinder(reader, rects)
            reader.End()            
        element = reader.Next()


def MarkImage(doc, page, writer, builder, rect, font, font_size, text):
    element = builder.CreateTextBegin(font, font_size)
    writer.WriteElement(element)
    element = builder.CreateTextRun(text)
    element.SetTextMatrix(1, 0, 0, 1, rect.x1, rect.y1)

    writer.WriteElement(element)
    writer.WriteElement(builder.CreateTextEnd())

def main():
    PDFNet.Initialize()
   
    doc = PDFDoc()
   
    eb = ElementBuilder()
    writer = ElementWriter()

   
    page = doc.PageCreate(Rect(0, 0, 612, 794))
   
    writer.Begin(page)  # begin writing to the page

    img = Image.Create(doc.GetSDFDoc(), input_path + "peppers.jpg")
   
    element = eb.CreateImage(img, Matrix2D(img.GetImageWidth()/2, -145, 20, img.GetImageHeight()/2, 200, 150))

    writer.WritePlacedElement(element)
   
    gstate = element.GetGState()    # use the same image (just change its matrix)
    gstate.SetTransform(200, 0, 0, 300, 50, 450)
    writer.WritePlacedElement(element)

   

    writer.WritePlacedElement(eb.CreateImage(img, 300, 600, 200, -150))
   
    writer.End()    # save changes to the current page
    doc.PagePushBack(page)

    # End page ------------------------------------


    doc.Save((output_path + "element_builder_00.pdf"), SDFDoc.e_remove_unused)
    print 'saved ' +  "element_builder_00.pdf"

    # sample document generated
    # find image bboxes

    rects = []
    reader = ElementReader()

    reader.Begin(page)
    ImageFinder(reader, rects)
    reader.End()
   
    print 'found ' + str(len(rects)) + ' images'

    # write the text
   
    font = Font.Create(doc.GetSDFDoc(), Font.e_times_roman)

    font_size = 12
    writer = ElementWriter()
    builder = ElementBuilder()
    writer.Begin(page, ElementWriter.e_overlay)
   
    for r in rects:
        MarkImage(doc, page, writer, builder, r, font, font_size, "Hello World")

    writer.End()

    doc.Save((output_path + "element_builder_01.pdf"), SDFDoc.e_remove_unused)
    print 'saved ' +  "element_builder_01.pdf"
   
if __name__ == '__main__':

    main()