I want to start playing around with computer vision. I am having a hard time deciphering the few available APIs and tutorials because they go into so much detail I really don't require. They talk about writing drivers which is not what I want to do. I just want, for the time being, to know what is the fastest way (shortest way) to just grab an image from a web cam without the extra error handling. In other words I just want to know what are the least steps needed to just capture an image and save it to a file. I am programming in C (in Ubuntu). Hopefully this will get me started with V4L. Thank you in advance.

Have a look at [url] http://www.linuxtv.org/downloads/video4linux/API/V4L2_API/spec-single/v4l2.html[/url] This is the documentation of how to use the V4L api, rather than how to write the device drivers. Appendix B has a sample application that opens the device and grabs some video from it.