If the image is raster artwork, such as JPEG, PNG, or GIF, it basically consists of coloured squares, pixels. The number of pixels determine how large you can make it without its losing clarity because the resolution becomes too low. With Illy, you may create vector artwork which can be enlarged freely without losing clarity, and the concept of resolution only applies when converted to raster.
Still, an image would make it easier to avoid saying something silly.
If Smart Deblur is out of the question (see post #3 in this Photoshop thread),
the best way will be to increase the size with all the blur, lock it, and draw on top of it with the Pen Tool (see instructions in the Helpfile).
You can use an online tool called Vector Magic (www.vectormagic.com) which basically takes any bitmap image and converts it into a vector graphic using points. You may have to do some adjusting but it's a quick and easy to do the conversion. Or, you could open the image in Illustrator, lock that layer, create a new layer and trace over it with the pen tool. That's the cleanest way but can take a lot of time depending on how comfortable you are with the pen tool.
If all you have is this file, then forget about any autotrace functions...
Correct. To elaborate:
There is such a huge misconception about autotracing and "resizing with clarity" as you put it (i.e.; resolution independence). There is no magic. You don't get something-for-nothing. Running an autotrace routine does not create additional resolution. And garbage in, garbage out still applies.
Entropy still rules. As in anything else, information is actually lost in mere "conversions" (especially automatic ones).True "resolution" is information. The amount of information (desired detail) and noise (undesireable detail) is just as fixed in a vector-based graphic as it is in a raster-based image. When you autotrace an ugly (in terms of low information and noise) raster image, you do not gain the ability to enlarge it "with clarity." You simply trade one kind of low-information ugliness for another.
Consider: What would be the "most accurate" automatic vector-based tracing of your image? It would be one vector square path for each square pixel in the image. That would be a "perfect" autotrace. And functionally it would be exactly equal to the original raster image. Absolutely nothing would be gained.
That's the big misconception about autotracing. Only by applying intelligence to reinterpret the image as a vector-based graphic will yield the improvement you seek. And autotrace routines contain no reall shape-recongnition intelligence. An autotrace routine, for example doesn't know "monkey tail" from "monkey eyeball". Only human intelligence does.
Learn to use the drawing tools. That's what vector drawing programs are really all about. There is no automagic shortcut.
Sorry for the mass problems. I didn't relize that this was a low quality picture. It was something I got of a website. I'm trying to dress up some switch plates for my grandson with a theme. What type of resolution should I be looking for.
Here is what I'm doing. I usually use fabric and sometimes paper to cover an electric switchplate. I have a process to do this. I'm going to try photopaper and do the same thing. Just trying to get a monkey theme for his room. I will go back and google for better quality pictures, once I get a response on the resolution or (I guess pixels?) that I should look for.
I don't think there are any copyrights apply to this picture. This is one small picture off of a wallpaper sample online. I have seen this design on other mfg wall paper. Also, it is not used for a commercial situation, not sure if this applies or not.
Sorry for the mass problems
There is no "mass problem." Your post is a frequently-repeated question and, as explained, is another instance of a very widespread misconception among newcomers to vector drawing programs about autotrace features (largely propigated by marketing hype). This being the Illustrator forum, unless otherwise stated, it is generally assumed you are talking from the perspective of commercial reproduction.
My also oft-repeated explanation still stands, as does the solution. It's not really a question of 'what's enough raster resolution for a good vector autotrace?' If the raster image is of sufficient resolution to result in an autotrace that looks good enough at the largest intended size, in all likelihood, the raster image itself would be of sufficient resolution to look good enough at that size. Othewise, you just trade "vector ugliness" (spikes, irregular jagged, or inaccurately shaped paths) for "raster ugliness" (pixelated undersampling). The resolution-independence advantage of vector graphics stems from the vector paths having been well-drawn to describe the meaning of the shapes they represent.
So if you really do need the resolution independence advantages of vector graphics, then you need a vector graphic intentionally drawn for that purpose. You either draw it or buy it.
You might try BigStockPhoto.com. Key "vector monkey" into the search. You'll likely find something suitable there for just a few dollars. Then you'll be using it appropriately, and there will be no question about legality. (You're really not supposed to just grab images from websites and use them without permission.)