How did the quest to make computers 'see' begin with simple line drawings and summer undergraduate projects?

The answer

Computer vision as an academic field traces to the 1966 MIT Summer Vision Project, where Marvin Minsky and Seymour Papert assigned a group of undergraduates the task of getting a computer to describe what it saw through a camera - expecting it to be a summer's work. The project made little progress, as the problem proved vastly harder than anticipated. Early breakthroughs came from David Marr's computational theory of vision at MIT in the late 1970s, which proposed that vision proceeds through stages: from a raw 'primal sketch' of edges and boundaries, to a 2.5D sketch capturing surfaces and depth, to a full 3D model. Larry Roberts' 1963 PhD thesis at MIT on extracting 3D geometric information from 2D photographs of polyhedral objects is widely regarded as the founding work of computer vision.