I am reading a research paper by "Andrea Censi" and "Davide Scaramuzza" on Low Latency Event Based Visual Odometry.
In the paper it is mentioned:
"Let the data from the camera be a sequence of tuples $\langle t_{i},y_{i}\rangle$, where $t_i$ is the timestamp, $y_i: \mathbb{S}^{2}\rightarrow \mathbb{R}$ is the image (here defined as a function on the visual sphere)."
I want to understand what this visual sphere $\mathbb{S}^2$ space is? Without this knowledge, the above said line and in-turn $y_i$ doesn't makes sense to me.
Kindly help.