The differences at the photoreceptor level have been addressed by others. The mechanical restrictions of the visual system were shortly hinted at by @gilhad andet al., but deserve more attention in my opinion.
First off, in darkness we cannot focus on an object and our eyes will move. And even when we focus on a specific point there is always movement of the eyes due to due to tremor, drift and microsaccades. Microsaccades are involuntary small movements of the eye (Fig. 1) that have received quite some attention lately. It is estimated they occur 1 - 2 times per second and they can reach amplitudes of up to 1 degrees of field of view (Martinez-Conde et al., 2013) and last for about 15 ms (Cui et al., 2009). It is thought that these movements prevent adaptation at the retinal level, and prevent image fading. Hence, images on the retina are constantly mechanically refreshed. The brain in turn stabilizes the image by correcting the image at the perceptual level through oculomotor feedback (Martinez-Conde et al., 2013).
Fig. 1. Microsaccades recorded by an eye tracker. Source: Martinez-Conde et al. (2013)
While a camera must be fixated on a tripod stand to allow for overexposure, our eyes cannot be fixated to the same extent, even when we try. Hence, combining exposures as indicated in the question is impossible and results in image blur. Instead, retinal images are constantly refreshed and when lighting conditions are too dim we cannot integrate photon input in the temporal domain.
Note, however, that photoreceptors do integrate photon input to some extent, given that higher luminance results in brighter perceptions. However, this operates only in the order of milliseconds and doesn't allow for long-term exposures as necessary to obtain images like the one shown in the great answer from @anongoodnurse.
References
- Cui et al., Vis Res (2009); 49(2): 228–36
- Martinez-Conde et al., Nature Reviews Neurosci (2013); 14: 83-96