Spherical detection : Color-based vs. shape-based and illumination problem

Image via Wikipedia

Forgive the technical title. I have to get this off my chest somehow.
It' s really bothering me when every single thing that I read carefully beforehand, arrive to a bitter conclusion. It doesn't work. Seriously, I don't want to frustrate myself so I am reading the same material over and over again in case I miss something in between. I began to worry whether my graduation project will eat me up in and out.
The subject is easy, but occlusion and illumination problem is overwhelming. Occlusion problem somehow can be solved with stereo vision but to be invariantly illuminated at the same time is almost impossible. I read hundreds of journal papers which only point to one direction. I only have minimum understanding for what they were talking about. The equations are mind-bogglingly scary. The related ones are even scarier. I began to wonder whether I am really a degree holder?
I guess this is the ultimate trauma. Seeing something that you firmly believe correct in the first place, only after months of immersing yourself in that belief; the truth is it never was. How sad. And I can safely say, not many people wanted to break-off their escalating commitment. Neither do I. Should I go and scream somewhere?
Thank God I can finally say it out loud.
Really.