I have a problem and do nor know how to target it. I have to estimate the quality of an image. This sounds easy, but in fact I only can work with the resolution. What I cannot estimate is the "information density" within that image (e.g if you resize a 40x30 pixel image to 4000x3000 pixel and then process this big image, this would mean that even a print of 26.67 x 20 inch would lead to a cool quality of 150 DPI, but in fact it will be a bad result...
What is a possible approach? Is there a way to determine the amount of artefacts or a parameter for "picture information density" within an image (when knowing the size the image will be printed)?
I need that to give customers a better quality odometer result on our website http://www.posterjack.com/. Actually I only calculate the DPI's and depending on the result we have different categories from "not recommended" to "excellent". Many customers are worried, since they get a good result and a bad poster... due to the hard compressing routines on cheap digital camera models. They have lots of Megapixels, but they also store tons of images on a small SD card...