Lighting is the critical—and often underappreciated—enabler of machine vision in agriculture. This presentation surveys the illumination techniques reshaping operations from the field to the packhouse: from visible-spectrum sorting to SWIR imaging that reveals defects invisible to the human eye, to multispectral approaches that unlock information about chemical composition and product quality. We also examine the unique challenges emerging in field-based systems like robotic harvesters—where bright sunlight and variable ambient conditions complicate reliable imaging—and how falling sensor costs are accelerating adoption across the industry.