The trap that a lot of garage inventors fall into is the idea that they have an absolutely original idea that will make them millionaires and that they must protect their idea and develop it in secret because otherwise it will be stolen immediately and exploited (or suppressed) by evil large corporations.
The sobering historical reality is that there are very few absolutely original inventions. Newton's comment about "standing on the shoulders of giants" is very apt, since most inventions and discoveries are incremental and evolutionary rather than revolutionary. I've also decided that there is a certain inevitability about some discoveries and inventions - once the right mix of foundational technologies exist and the right ideas are floating around, certain new applications and combinations of them are almost inevitable and simultaneous independent invention is likely. Many ideas I've come up with are like that - I come up with ideas that are original to me, but with a little research, I find that it's already been done or that someone else has just announced they are working on the same idea. You just have to deal with that as a reality and keep trying.
The second part of the trap is finding that the idea has been explored before, but was never realized into a product or that it was realized into a product, but went nowhere. There are many causes for this. Sometimes the idea sounds valid, but turns out to be impossible (e.g., free energy, perpetual motion, or inertia-less propulsion), sometimes the idea and product works perfectly, but the market just isn't there - either the target market is too small, it doesn't appeal to many for some reason (fashion, culture, etc.). Sometimes the idea was implemented badly causing the market to reject it. In some cases, this is a great opportunity to improve it. In other cases, public reaction was so bad that the market will reject similar items for a long time. Look at how long it took PDAs to become accepted and all of the backlash that early models received.
Take, as an example, this Arduino forum. You read one post on one of the subgroups where someone does something cool. Then you read another unrelated thread where someone else is doing something cool. Then you get a bright idea - "I can combine aspects of both of those ideas into a single gadget!" Congratulations, you just had a totally original idea for an invention. Unfortunately, chances are, dozens of other readers also just had the same original idea. Even then, all of you are a long, long way from a marketable product.
So, for 99% of ideas for inventions, you need to be honest with yourself and be willing to let go of ideas. Often, design, fashion, and usability are as, or more, important than functionality. Look at the iPod, iPhone and other Apple products as an example. Each of these devices offer only incremental technological innovation, but their ease-of-use, "cool factors", and design elements make them a hit. Read books on design, such as "The Design of Everyday Things" or "The Evolution of Useful Objects" for a better understanding of how design is as important as innovation and how having the groundbreakingly original idea doesn't always mean you win the market - it may be the 5th evolutionary generation of an idea that hits it big.
Finally, as others have said very well, if you are still convinced that you have an original idea that you feel is marketable, you still have a lot of work to do. Business plans, multiple prototypes, usability testing, market research, pitches, and so on. You don't want the first investor you approach to say, "I had something like that ten years ago and I hated it!" If something similar was available then, you have to be prepared to explain why you have fixed all of the problems and have a plan for telling the market why your idea is better. If it is completely revolutionary, you may have to build the demand through marketing - remember that there was a time when people didn't think they would ever want or need a cell phone, for example.