Converting Pixels to Vectors Using Image Trace in Adobe Illustrator

Use Illustrator’s Image Trace to convert high- or low-resolution pixel-based images into editable vector graphics, adjusting settings like color count, path fitting, corners, and noise.

Convert pixel-based images into editable vector graphics using Illustrator's Image Trace feature for clearer, scalable artwork. Understand how input image resolution, tracing settings, and cleanup options determine the quality of your final vector output.

Key Insights

  • Image resolution significantly impacts output quality when using Illustrator’s Image Trace; high-resolution images produce more accurate and detailed vectors, while low-resolution inputs lead to less precise results.
  • The Image Trace panel provides advanced controls such as path fitting, corner emphasis, and noise filtering, allowing users to fine-tune the vectorization process based on the nature of the artwork.
  • This course demonstrates that expanding traced images is essential to convert them into editable vector paths, enabling further customization such as removing unwanted backgrounds or modifying anchor points.

This lesson is a preview from our Illustrator Certification Course Online (includes software & exam). Enroll in this course for detailed lessons, live instructor support, and project-based training.

This is a lesson preview only. For the full lesson, purchase the course here.

Illustrator has an image trace option, which allows you to convert pixel-based art into vector art. So instead of having to manually trace very detailed, complicated art to create vectors, we can use this image trace feature. Let's take a look at how it works.

I'm going to create a new file. This is going to be my letter-sized page, wide aspect ratio, and using inches. Into this, I'm going to place my pixel-based art.

So if I want to make edits, change colors, print this at infinite resolution like we do with vectors, I can import pixel-based art and convert them into vectors using the image trace feature. I need to place or import that artwork. They call it place, but this is an import where I can bring in a pixel-based graphic.

Now, resolution does matter. Resolution is how many pixels per inch we have. Lower resolution images have less detail.

Higher resolution images have more detail. I'm going to show you tracing both low-res and high-res to let us see the quality of the output that we should expect. So I'm going to start by bringing in the low-res, placing that in.

I'm going to click to place it, and I can make a duplicate over here. So let's say I size this and I place it in here. Now, first of all, as I'm resizing, I need to hold shift if I'm going to change the size of this, but let's say I choose a certain size for this, and then I want to replace this with the same image.

It's just actually a higher resolution version of it. I'm going to duplicate this by option dragging or alt dragging on Windows, and anytime you have one pixel-based graphic here, you can actually replace it with a different pixel-based graphic. So these are essentially images that I brought in, and when I zoom in closer, we can see the pixels that make this up.

So this is pixel-based graphics. Pixel-based graphics have these little square blocks or pixels that make it up, and I want this to be vector graphics, not only for editing but also for the clarity and crispness of its edges. So I can select the image that I want to replace and go to place a new image, choose the high-res version, and before I hit place, I want to make sure that I choose to replace.

This will replace the selected graphic. And now, if I zoom in on this one, notice how much crisper this one is. Yes, it is still made of pixels, but notice how much more detail there is here than there is in this one.

This is highly pixelated, and over here, the same kind of curve. Yes, there are pixels, but there are a lot more pixels. That is higher resolution.

We have more pixels per inch. So let's see what the quality difference is by trying to trace a low-res version and a high-res version, because one of the challenges is we often do get low-res kind of graphics or low-res pixel-based images, and we want to trace them to convert them into vectors, thinking that we're going to get a good result, but you are limited by what you bring in. The worse you bring something in, the worse you're going to get out.

The better your input, the better your output. But let's take a look and see what the quality is between these two. We can select this graphic here, and over in the properties panel, I can click image trace.

I'm going to choose three colors because that's all I need: white, black, and orange. I'm going to do the same thing over here. Choose image trace, three colors.

This one's going to take a bit longer to do, so if your computer pauses for a moment, just wait until it's finished, and let's zoom in and take a look. First of all, you will notice how crisp and clean everything is, but look at the quality or lack thereof, unfortunately, because there wasn't a lot of detail to begin with, so it's kind of not the best output here, especially these kind of letter forms here. Not a great thing.

This is definitely much better. When you look at the shapes, it's definitely better quality here. Is it perfect? No, but here, for example, I can see the registered trademark here.

It just looks like a blob. Now, if I want to see kind of the before and the after, I can select this, and next to preset here, I can click this button to see the image trace panel. I can also open that image trace panel up by going to window and going down to image trace, and here, if I want to see the tracing result, which I'm seeing right now, versus the before, when I press and release, press and release, so the before, which is the pixel-based graphic, you can see I can't really even read what that registered mark is.

I know conceptually what that is, but I can't really see it, and so Illustrator is kind of getting confused. Now, if I come over here and I press and release, see, that's a much better quality input, so I'm getting a better quality output. Now, things are not always perfect, and I might need to tweak it, and maybe I can get something better, so zooming in here, there are some things that I can change.

For example, here under advanced, which might be collapsed by default, I can open that up, and I can choose the path fitting, so a tighter fit with high or a looser fit with low, so if I go to low and give it a moment to retrace it, or if I go to high and give it a moment to retrace it, in theory, you might think, well, I always want a tight fit, but when we think about converting pixels, pixels are inherently imperfect, and so tracing those really accurately actually sometimes shows you details that you don't want to see, because it's kind of tracing the pixels versus going for the intent of those. Now, on the other hand, if you are too loose about things, maybe sometimes it's not as accurate as you'd like to be, so it depends on the graphic as to where you want to be, probably somewhere in the middle, but you'll have to play with that. Corners, it can also have a balance of, do you want fewer corners where things are more rounded, or do you want more corners? Again, you can go too far and have too many corners, but when you're getting those kinds of points right here, is that going to be more of a rounded appearance, or is that going to be more of a corner point appearance? Also, the other thing you can do is look down here and see how many points it's creating, so these are how many anchor points, how many paths, or kind of segments of lines it's creating, and if I go to a high fit, notice we get more points.

Now, also more points, once we expand this, will be more places to edit, which means more complications to edit, so there is a balance of the fewer points I have, the easier it is to edit that, and also the smoother maybe that final result is. Noise is going to be considered little specs or things that might be traceable. Now, this particular original image, if I press and hold to view the original, this doesn't really have a lot of dirt or noise to it, so it's not really tracing a bunch of noise, and so when I look at that, this is not the best to explain noise, but the idea is that if you have it trace every little teeny thing, every little spec of dirt, every little teeny piece of information, it might trace lots of little things.

If you tell it to kind of ignore smaller noise, then things have to be at least 34 pixels in size to be considered traceable; little things that might be traced will be ignored. Lower that down, and like I said, in this case, not really much of a difference here because we don't have those pieces of noise to really worry about in this case. When you're all done, and you want vectors, we do have to expand this, which means to render the vectors.

If I go up to view and outline, notice we don't actually have true outlines, so when I go back to preview, these are still pixel-based graphics that are being vectorized, and I'm seeing the preview, but I haven't actually finished the job, so I have to click expand, and now this is no longer a traced object. This is now vectors. If I go into outline view, you can see that those are actual outlines, and those are things that I can go in and change.

I can go in with my direct selection tool, choose my points, move them around, and edit those things. Go back to preview mode, and so anything that is not perfect, I can go back in and make changes to. So this is one thing that we can work with, and hopefully, as you've seen here, the better quality you bring in, the better quality you'll get out.

Let me show you one other image. It's more of an illustration, and I'm going to place this in. I've got this sketch.

I'm going to click to place it. Now, this could be something that you could hand sketch in. This is a very refined sketch, or this could be a piece of digital art.

So whether it's a hand sketch, a painting, or even a photo, you can bring in photos. You can bring in all sorts of things and then convert them to vectors and see how good a job it does. Now, things that are photos, kind of continuous tone, like gradations, they're not going to work as well.

You really want things that are more flat, kind of pieces of art, rather than photographs. Photographs are best represented by pixels, not vectors. So here I'm going to select this, and I'm going to image trace it.

So I'm going to come over here and choose image trace, and choose a preset. Now three colors are not enough. I've got multiple oranges here.

I've got the gray, so I need more colors. So let's say I go to maybe like 16 colors. I might be able to get fewer colors, but we'll start with 16.

They're saying that this is a larger, higher-res image. Do you want to continue? Yes, in fact I do. So click OK.

This takes a little bit longer to render because there are a lot more vectors to draw. And zooming in closer here, this has done a very good job. Now there are imperfections, things that I would want to ultimately smooth out and correct.

And if I bring up the settings over here, open up the image trace panel, just like I did before, I can preview the original by pressing. And then when I release, I see the after. Press to see the source image, release to see the after.

And this way I can see how good or bad a job it's done. It's actually done a pretty darn good job of doing the tracing. Again, we can play with the options here, whether there are smoother curves or more corners. So here there are some more angular parts.

I would rather this be a bit smoother, so I think I will go with fewer corners. If you think you can get away with fewer colors, then right now it's actually, even though we said we can do up to 16, we've actually chosen 14 colors. But I might be able to reduce it.

I might not need quite all of those colors. So let's say I go with less. Now I'm going to go with too few colors.

I'm going to get down to like, let's say, six colors, and you're going to see what it looks like. And then I'm going to bump it up to get more colors. So notice now here, see how I've lost the shading that's there at the eyes? Because I went with too few colors.

So I'm going to pop up a couple of colors. See if that's enough. Nope, still not enough of the shading.

So I'm going to pop up a few more colors. There we go. That is better.

Now, if you want more colors, of course, you can still go up with more colors. That'll give you more shapes and more differentiations between different colors in different areas. All right.

So let's say we are happy with that result. I can click expand. And then this does have a background, which maybe I don't want.

So when I click on this right now, this is grouped together. I can double-click to go into that group, select the background color, and delete it. Now I can double-click to go out.

And I've now got this isolated on a background of whatever I choose. I can create any sort of shape background. And I'm going to send this back.

Arrange, send it back. And now this is sitting on top of that. So if I want to put it on a circle, for example, I can draw an ellipse, holding shift, and send it to the back.

And now I have that sitting on a circle, kind of breaking out like so. 

Dan Rodney

Dan Rodney has been a designer and web developer for over 20 years. He creates coursework for Noble Desktop and teaches classes. In his spare time Dan also writes scripts for InDesign (Make Book Jacket, Proper Fraction Pro, and more). Dan teaches just about anything web, video, or print related: HTML, CSS, JavaScript, Figma, After Effects, Premiere Pro, Photoshop, Illustrator, InDesign, and more.

More articles by Dan Rodney

How to Learn Illustrator

Build practical, career-focused Illustrator skills through hands-on training designed for beginners and professionals alike. Learn fundamental tools and workflows that prepare you for real-world projects or industry certification.