In 1992 commercial satellite remote sensing was a $190 million industry and aerial remote sensing was about $2 billion (CSTS, section 3.2.1, page 70). The main reason that aerial remote sensing is a larger market is that it can provide 1 meter resolution (various satellites on the drawing board can do this too) (CSTS, section 220.127.116.11, page 74). The demand would be approximately 6 times higher at $4 million per launch than at $14 million per launch (the current price); launch costs below $4 million would not increase demand because ground segment costs would dominate (CSTS, section 18.104.22.168, page 84).
Categorizing the providers is not always easy, but roughly speaking the players for the space segment (with emphasis on the commercial) are:
Their focus is now QuickBird:
Their first program, EarlyBird, was cancelled in 1998 following the failure of the EarlyBird-1 satellite in December 1997:
There are a large number of others, which I won't try to list, who resell images or offer value added services such as image interpretation. In fact the value-add segment of the market is larger than the space segment (CSTS, section 3.3.2, page 71). For an overview of the value-add industry see "Geographic Information System", Trends in Commercial Space - 1996, Keith Calhoun-Senghor, Editor, Office of Air and Space Commercialization, U.S. Department of Commerce.
One nice page is from the Federation of American Scientists. The page mainly contains regulations and discussion of policy issues (for example, the US government might restrict images of Israel).
Remotesensing.org is an interesting site, which focuses on free software for remote sensing applications. Relevant mainly because this kind of thing has the potential to change the way the value-add market does things.