The rodent visual system has attracted great interest in recent years, owing to its experimental tractability, but the fundamental mechanisms used by the mouse to represent the visual world remain unclear. One key step in representing the world in our visual system is that of perceptual organization, whereby individual objects are identified and segmented. An outstanding question is the extent to which perceptual organization is a common computational principle across mammalian visual systems. To answer this we looked for neural and behavioral evidence of such a process in the rodent. We find that mice can segment objects using contrast cues but fail to utilize motion cues. This is in contrast to primates who can use either cue. Lastly, we model the representations in the rodent visual cortex and find they are consistent with a feedforward cascade of computations. This work highlights both the strength and limits of the rodent as a model system for mid level vision.