Esem has the right idea, but the requests for pages is still not truly indicative of “unique visitors.”
A visit to a page with 3 embedded graphics will generate at least 4 “requests” - one for the page, and one for each of the three graphics.
That said, the number of pages requested does not really indicate how many “unique visitors” have visited the page, because, without special programming (cookies, session, etc) the “stateless” nature of the web means you can’t differentiate between a single user visiting a page five times, or five visitors requesting the page one time.
You can get some indication of this by checking the IP address of the visitor making the request, but that is not completely reliable, as many (most) of your visitors will be coming to your site using a dynamic IP address assigned by their ISP (and it will re-assigned to subsequent users when no longer needed by a given user to whom it had been assigned).
Without programming to retain the “state” of a visitor, you can only guess as to the number of unique visitors, and your best tool for making this “guess” is the number of unique IP addresses visiting. Unlike page views, or requests, (both of which will always be higher than the actual number of unique visitors - page views being “higher” and requsts beng “much higher”), this number may be lower than the actual number of unique visitors because of re-used IP addresses (though it will also cause you to count robots and search engine crawlers as “visitors”)
Does that help explain it at all?