Er, how does this answer the poster's question?
The problem with any answer to the OP's question beyond that of "this question takes a bunch of differing concepts and mashes them together in a nonsensical way" is that such an answer would have to deal with the problematic presentation of said concepts.
Also, this is not the first time this kind of problematic question has been asked:
A Problem with QED equation E = h f
For the emission and the absorption of photons by atoms, QED relates the energies of photons to their frequencies by E = h f equation, and apparently in these cases the amplitude of the photons are irrelevant, but for the photons that are generated by the electronic circuitry the issue of photons’ amplitude should be considered differently, because Equation E = h f fails to differentiate the energy differences in AM radio signals and their carrier signals, so photon’s energy has to be related to its amplitude.
This is baffling and perplexing, how could QED theory that it is empowering our amazing technologies harboring such flaws at its core foundation?
The dual characteristic of photon has made its structure so ambiguous and so mysterious, that nobody understands its structure, isn’t it odd that its energy is defined with such a trivial equation?
I tried to answer, in some detail, the problems with the presentation of QED and related physical theories and concepts in that thread. Here, I did something similar but in less detail.
Basically, the problem here is again (as was true in the previous thread) that some mostly outdated theory, framework, or approach from physics or engineering such as classical optics (in this case) or radio waves (in the previous case) is used to "uncover" some problem with the nature of quantum electrodynamics.
The first issue is that one does not try to critique a theory like QED or even classical electromagnetism by using out-dated "laws" and descriptions from optics (or radio waves), however useful they may be in application today as approximations.
The second, deeper issue is that QED is not simply quantum mechanics applied to photons and light. It is drastically different from QM in structure, mathematical nature, sophistication, etc. One cannot simply take for granted popular presentations of e.g., the uncertainty principle as it is to be found in texts on QM and assume that either the conceptual structure or the theoretical structure of the uncertainty principle remains in QED. It doesn't.
Basically, this question:
For the mirror reflection if each photon of the incoming image on the surface of the mirror, it has to go through the uncertainty in the direction of the emission, and also it has to deal with the uncertainty in the time of the emission, as it is discovered by QED
doesn't make sense to begin with because there is nothing like what is described above that was "discovered" or that exists in QED. That's without the problem of dealing with the fact that in quantum optics and other areas of physics in which semiclassical, quantum, and other non-relativistic approaches to photons in which it makes sense to ask questions about single photons, this isn't a sensical question.
The uncertainty principle in QM is not something that anything goes "through"; rather, it is due to the non-vanishing of the commutator of observables in the operator algebra(s) of QM. It is thoroughly quantitative.
If you wish to understand what would happen in a given scenario QM, QED, optics, etc., you can in general find the answer (or at least whether the question makes any sense) by
actually writing down the theory or theories in question and then putting in the appropriate terms for the situation. This is a major advantage in physics: theories are not merely frameworks as they are in the sciences in general but tend to be mathematical equations or sets of equations. You can write down QED (and indeed the standard model) rather easily and literally point to the relevant portion of the Lagrangian (or, if you want to get
really simplistic, you can eschew equations entirely and draw the relevant diagrams using the appropriate Feynman rules).
It is difficult, in this case, to do so not because QED fails, but because the question starts and continues in a manner that should make any attempt to actually formulate the question in a sensible manner immediately render clear its futility. The question continues in a similar manner:
What happens at the surface of the mirrors that our well-tested uncertainty rule of QED for the reflected photons totally fails, and instead, we have the certainty of the Optic Law for the reflected photons?
There is no such uncertainty rule in QED and there is no optic law that can simultaneously tell us what happens to a photon in a physical situation that also includes the interaction of the photon with a macroscopic instrument (which, even in the simplest quantum mechanical situation, means that special kind of entanglement we normally just call "measurement" or even "collapse of the wave-function" if we wish to exclude the macroscopic device we cause to interact with the quantum system). The entire question is posed in such a way that a naive treatment of photons given in popular accounts of QM is said to hold for QED when it doesn't and them this treatment is supposed to somehow be put into question because of some "optic law" which would fail to account for photons in the first place, let alone how or if we can make sense of a single photon interacting with a mirror in the manner described in the question. Then we are told:
QED All-Path method and Wavefunction method are not supposed to implement the detection point which it is dictated by the Optic Law, because QED tells us “the possible detection points” for almost all the emitted photons are very different from “the detection point” that it is offered by the Optic Law.
Here the problem changes somewhat. Now we are some what beyond simply the issues involved in trying to answer a question which confuses a number of concepts and theories and is fundamentally ill-posed. We are now given problematic descriptions of these concepts and how "the Optic Law" should give us pause here. I honestly am not quite sure what the first assertion is intended to refer to when it conflates Feynman integrals and wavefunctions into some method that has some implementation rule concerning detection points and the Optic Law. I DO know that there is no method in physics as described and the OP is confused about something when it comes to detection points and this "optic law" notion.