Vision: eyes vs cameras I think this is related to physics, and light waves.
Is there some materials that can be seen by human eyes and can't be seen by cameras? (Imagine a special writing) Is that even possible therotically?
 A: Short answer: In the simplest physical sense, no (see bolded item below).
But it depends on what you mean by "seeing".
Note that human vision blurs the line between Physics and Biology - the brain is responsible for a lot of what we think of as "seeing".
Longer answer:
Since you tagged this with visible-light, I'll take the liberty of assuming those are the wavelengths you care about.  (As the commenters pointed out, cameras can be built to be sensitive to other wavelengths outside the visible range, but to me this seems outside the spirit of the question.  It's definitely not a 'fair' comparison to the human eye.)
Ultimately "seeing" has two components:

*

*Translating a spectrum visible wavelengths into a single (or in the case of color, three) brightness or intensity values

*Image processing the brightness/intensity to understand what was sensed in #1.

#2 is no longer purely physics - a 'brain' is now involved.
For #1, there is little difference between a human eye and a camera for visible wavelengths.  (Yes, some cameras can sense other wavelengths, but we will ignore that for now, per above.)  Both are sensitive to visible wavelengths, and both translate them down to one brightness or intensity value (three for color).  You can web search for 'wavelength sensitivity of cameras' to prove this to yourself, if you want.
Perhaps the biggest relevant difference is dynamic range - the human eye has much better dynamic range control than a typical camera.  (See this question from photo.stackexchange.com for info.) Cameras are fast approaching the eye's resolution (think pixel count) as well.
With wavelength filtering, apertures, ND filtering (etc), a camera could be created with almost identical wavelength and brightness response to a human eye. For all intents and purposes, the signal measured by such a camera is the same as measured by the human eye.  In that case, it doesn't matter what the material is - the human eye sensor or the camera sensor would register an equivalent measurement under the same conditions.  Even a typical camera without these extra properties responds similarly enough to a human eye; that's the point of the camera, after all.  :)
If you just wanted the answer to #1 (this being Physics), you can stop here.
For #2, now that we have established that input devices (human eye or image sensor) have basically equivalent responses, we now shift focus from the input devices to the processors that evaluate the data (brain or software algorithm).  Here is where we leave pure Physics.  That said, the brain is controlling the inputs that are relevant to #1. For example, a brain will dynamically adjusting aperture size based on signal input.  (Even eyes can overexpose.)
If there is some software trying to "see" your proposed material, it's probably trying to distinguish things inside its field of view from either a 'background', or from other things.  You could in principle design a material to confuse this algorithm, if you knew how the algorithm worked.
That said, human brains are highly adaptable - they can learn new algorithms.  If you allowed the brain of your camera to do this as well, then the answer is still no, as the same logic a human brain learned to identify a material could be encoded into an algorithm for the camera.
It's up to you how far you want to take this thought experiment, such as what limits you want to set up for the human (eye + brain) and/or the camera (sensor + software).  To me, the comparison should be 'pure' and 'fair', where the non-human system has similar specifications (resolution, dynamic range, even a brain system that can learn), so I would still answer "no".
