Recreating an image from a photometer or similar light-detecting device? I'm thinking if it is possible to recreate an image from data from this kind of device.
It is known analog signals theoretically have infinite resolution, but since we use them in discrete systems (bits, for example) they lose that behaviour.
My first assumption is that the question related devices aren't discrete enough to gather data to recreate the image, but I'm not good at Physics to know more than that.
 A: Some background.
You want to detect the image of an object. 
First, either 1) you illuminate it with some light source [a lamp, the sun] or 2) the object itself radiates light out [a star, a fluorescent sample]. Imagine to divide the object in many small parts (voxels): to reconstruct the image of the object, you need to detect independently the light coming from every one of them. 
There are different ways to do this. The most used is to use a broad illumination which illuminates all the object, a lens, and a pixelated detecting surface (a ccd camera), where each pixel is an independent detector. The lens will steer the light coming from different parts of the object onto different pixels. The intensity detected by all the pixels, when put on a screen in form of a matrix, shows the reconstructed image.
Therefore, a "single pixel detector" like the one you mention, cannot work in this (usual) configuration, because it lacks the independent pixels. 
There are other ways to reconstruct an image from a single detector, but it becomes complex and probably it's out of the scope of your question. You can use a smart illumination scheme together with data analysis (like here). Another scheme widely used in microscopy is the confocal microscope, which is used with fluorescent samples. 
