Images extracted are mirrored on x-axis.

[yashpundir]

Hi, I'm new to PDFplumber and I'm trying to draw the images detected by the library on the page itself. I do this using the Page.draw_rects() method and use the x0,y0,x1,y1 coordinates obtained from the Page.images properties. But the rects being plotted are mirrored on the x-axis. Am I providing the coordinates in a different order than what is expected by the draw_rects method?
Here is my code:
rotomecPDF = pdfplumber.open(rotomec) # load pdf
roto_page6 = rotomecPDF.pages[5] # select page 6
pg6_imgs = roto_page6.images # find images on pg6
roto_page6.to_image().draw_rects([(i['x0'],i['y0'],i['x1'],i['y1']) for i in pg6_imgs]) # draw a rect for each img

Here is the pdf
Rotomec_5002.pdf

And here is the output of the above code. As you can see the images are inverted on the x-axis.

How can I fix this. I know I can write a function to invert the points on the coordinate plane. But I think I'm doing something wrong as the chances of this being a bug in the library are really low.
Thank you in advance and apologies for my English.

[petermr]

When I display your post (in Chrome) and compare with page6 in the PDF (in Acrobat) the images are both in the same orientation.

[jsvine]

Hi @yashpundir, a couple of notes to help you on your pdfplumber journey:

• y0 and y1 are the vertical coordinates in the PDF's original coordinate space, where the origin is the bottom-left of the page; pdfplumber's methods, however, use the top-left of the page as the origin. So you'll want to use the top (distance of the top edge of the object from the top of the page) and bottom (distance of the bottom edge of the object from the top of the page) attributes instead. I.e., (x0, top, x1, bottom).
• PageImage's drawing methods can accept entire graphical objects, e.g., im.draw_rect(page.images[0])

So you could do this:

import pdfplumber
pdf = pdfplumber.open("path/to/file.pdf")
page = pdf.pages[5]
im = page.to_image()
im.draw_rects(page.images)

... which returns this:

[bikallem]

Is there a way to extract the said image, i.e. page.images[0] into a png/bmp/jpg format?

[jsvine]

If your goal is just to extract the images in a PDF, you might have more luck with MuPDF, which includes a specific method for that: https://pymupdf.readthedocs.io/en/latest/document.html#Document.extract_image

pdfplumber does not currently have much direct support for image extraction, unfortunately.