Lets think about a CCD imager with 1290 horizontal by 1200 vertical pixels. that's about 2.3 million pixels. For now ignore the issue of color. There can not be 2 million wires coming out of the CCD chip. So what happens is the information is shifted out a row at a time. Now this still means 1200 wires coming out if you had one per row and that's a lot better than 2 million coming out if there had to be a wire per pixel. Of course there are schemes to reduce the number of wires even more.
So now we have this CCD imager and it's ready to take a picture. Think of each pixel as a little light bucket. It accumulates light. At some point in time, the bucket from the pixel transfers into a holding bucket that's right next to the pixel. Then there is a bucket brigade that takes the 1290 buckets in a row and shifts them out of the imager. Electronics outside of the imager records how much "light" is in each "bucket", and then that device (the computer controlling the camera) has all the digital values for the 2 million buckets - the digital picture.
When you transfer the accumulated light from the pixel bucket to the bucket brigade, the accumulated light bucket is empty and it starts to fill again. How long you give it to fill up is the equivalent of the shutter speed in a traditional (or is it old fashion by now? ) SLR film camera. The less light you have, the more time it takes for the bucket to get some light accumulated in it.
CCD stands for "Charge Couple Device". The light in the bucket is really a charge that's accumulated. The "couple" part of this refers to the ability to "shift" the charge down the line to get the charge out of the array.
Of course the devil is in the details, and there are many factors to play with including but not limited to the surface area of the pixel (the light bucket size), the size of the brigade bucket (determines how noisy the whole system is), the speed at which you transfer the buckets around (it takes time to empty a bucket), issues with color, issues of silicon having different sensitivity to light at different wavelengths (colors), infrared filtering, and so on.
The higher end digital SLRs have mechanical shutters so that other advantages can be had, such as high speed (no blur) pictures. The area and sensitivity of the CCD array play into the system design of the camera.
Lower end point and shoot do not need mechanical shutters, they use the "electronic shutters" by controlling the fill time and dump cadence of the pixel accumulating buckets and the bucket brigade.
Finally, please accept that this is a hand waving argument and there are a billion details I could be called on. I hope the general picture is clear :laughing:.
Pete