American intelligence could not get a clear photograph of Soviet territory before August 1960. The first successful Corona mission that month returned 1,680 images from orbit, program — formally known as photographic reconnaissance and one of several categories of intelligence satellite, alongside signals intelligence (SIGINT), electronic intelligence (ELINT), and early-warning systems — would go on to fly 145 successful missions over 12 years and return 800,000 images of denied territory.
Corona used a film-return system: the satellite would take photographs, expose film, and return a reentry capsule containing the film to Earth. The capsule was caught in mid-air by a C-130, or splash down in the Pacific and be recovered by a ship. The first twelve Corona launches failed; the first successful mission returned 1,680 images of the Soviet Union, China, and other targets. The program ran for 12 years, with 145 successful missions and 800,000 returned images. Corona was declassified in 1995.
The transition to digital imaging was the most important technical change. The KH-11, first launched in December 1976, used a CCD array to convert the optical image into a digital signal, transmitted to a ground station in near real time. The current U.S. reconnaissance system is built on KH-11 derivatives and remains largely classified.
The Soviet Union operated the Zenit and Yantar film-return and digital series. The Soviet program was less productive than the American one, limited by the smaller Soviet launch vehicles and a less reliable film-return system. The Soviet Union relied more heavily on signals intelligence and on agent networks, which is one of the reasons it overestimated American military capabilities during the 1970s.
For the rest of the military space article and the space race overview.