The visual system begins with photoreceptors (rods and cones) in the retina that respond to light wavelength (cones) and intensity (rods). Retinal circuits extract local contrast and motion features before sending information to the brain. The lateral geniculate nucleus relays information to visual cortex where neurons are organized retinotopically (neighboring cortical neurons respond to neighboring visual field locations). V1 simple cells detect oriented edges; V2 and beyond process increasing complexity of features (curvature, color, motion, faces).
Examine retinal structure and rod/cone distribution across the retina. Study receptive field properties of retinal and cortical neurons. Trace anatomical projections from retina through LGN to cortical areas. Map visual field representations in cortex.
The eye is a camera / all visual information enters consciousness / V1 is the only visual area / color is processed only in cones.
You already know that sensory transduction converts physical energy into neural signals, and from your study of photoreceptors you know that rods are sensitive to low light intensities while cones (concentrated in the fovea) mediate color vision and fine spatial detail. The important insight here is that the retina is not a passive camera sensor — it is an active preprocessing station that performs significant computation before signals ever leave the eye.
The key structure enabling this preprocessing is the center-surround receptive field of retinal ganglion cells. Retinal circuits wire photoreceptors through bipolar and horizontal cells such that each ganglion cell is excited by light in a small central region and inhibited by light in a surrounding annulus (or vice versa). This arrangement makes ganglion cells maximally sensitive to local contrast rather than absolute light levels — they fire vigorously at edges (where brightness shifts abruptly) and are relatively indifferent to uniform illumination. This is why you can read in a dimly lit room: your visual system extracts contrast structure, not raw brightness. The retina thus sends a compressed, edge-emphasized representation of the visual scene down the optic nerve.
Signals from the two optic nerves partially cross at the optic chiasm — fibers from the nasal retina (carrying temporal visual field information) cross to the opposite hemisphere, while temporal retina fibers stay ipsilateral. The result is that everything in your left visual field is processed by your right hemisphere and vice versa. Signals then relay through the lateral geniculate nucleus (LGN) of the thalamus, which is organized into six layers: the two magnocellular (M) layers carry motion and coarse spatial information; the four parvocellular (P) layers carry color and fine detail. This segregation is maintained into cortex.
In primary visual cortex (V1), neurons respond to oriented edges — a breakthrough discovered by Hubel and Wiesel using moving light bars. Simple cells have elongated receptive fields that respond to edges at a specific orientation and location; complex cells are less position-specific but still orientation-selective. V1 is organized retinotopically: neighboring neurons respond to neighboring locations in the visual field, with the fovea overrepresented. Beyond V1, visual processing diverges into two major streams. The ventral stream (V1 → V2 → V4 → IT cortex) processes object identity — shape, color, faces — answering "what is it?" The dorsal stream (V1 → V2 → MT/V5 → parietal cortex) processes spatial location and motion — answering "where is it and how do I interact with it?" Damage to these streams selectively impairs different capacities: ventral stream damage produces visual agnosia (inability to recognize objects); dorsal stream damage produces optic ataxia (inability to guide actions to objects) despite intact recognition.