Describe the application, then it will be clear that you need to:
1) I'm doing the layout for which the manufacturer needs to produce a package (just b & W text and icons), send to the manufacturer in PDF
2) the Manufacturer loads it into their software, which machines print packages, converts it to your format and save the picture. Sometimes the manufacturer changes something in the layout.
3) then the manufacturer sends me a picture, as it is some will look like (exactly the same), I check the eyes and confirm or indicate that something is missing or wrong placed.
The designers of the manufacturer can sometimes delete important text or icon, with or without notice, to change the blocks.
Find out why they did it, for what purpose, and how to avoid it in the future impossible is Asia, where production of any level of sinning in this and information of them to pull the impossible, "so that they are all excellent", only to follow and to point out the shoals.
Of course sometimes you can't keep up. I would like to automate this process, using a differ image to upload your picture and screenshot clipping sent by the Manufacturer (usually it is sent a huge canvas with different pictures).
Saw online and neolayn differi images, but they can't ignore aspect ratio or white margins on the sides and either refuse to work or eventually show a hodgepodge of the two almost identical images, simply because of slight inconsistencies proportions
And something like "read the pictures from left to right, we see the same black spots of certain forms (letters and icons) in one and in another picture in about the same places, go ahead. And here there is no flaw, and here is, marked by a red dot".
There is something similar? Toll-free not important.
Or are there any other ways to solve the problem, "machine learning" there or something?
Thanks in advance!