From the course: Photoshop 2024 Essential Training

Removing objects using Content-Aware and Generative Fill - Photoshop Tutorial

From the course: Photoshop 2024 Essential Training

Start my 1-month free trial

Removing objects using Content-Aware and Generative Fill

“

- [Instructor] Photoshop's Content-Aware fill and generative fill technologies are an excellent way to remove objects in an image and fix areas that are distracting to the viewer. In this photograph, I want to remove the distracting pieces of ice from the left and the right hand side of this image. We can see in the layers panel, I have a background and a retouching layer that we created in the previous movie. I'll start by tapping the L key that will select the lasso tool, and then I'm just going to make a rough selection around the ice on the left hand side, and then I'll choose edit and then Content-Aware fill. This is a relatively small area and I think Content-Aware fill will have no problem with it, but why are we not seeing anything? So this is our original, this is the preview area. Well, it's because we don't have sample all layers turned on, so let's enable that and now we can see the original and then the preview. I'm going to use one of the three sampling options. The auto, the rectangular or custom. Here, auto works very well. We can see that ice has been removed. I can also use rectangular. That just makes a rectangular selection and we could use custom if we wanted to draw in our own area, but most of the time I find that auto does a much better job. There are additional options down here for color adaptation. The color looks very good to me, but if not, we could change that to maybe high or even very high, and that would enable Content-Aware fill to change the colors. I'm going to leave it set too high for now. We can also use rotation adaptation. I'll make that low and then that will enable Content-Aware fill to rotate the area in the image that it's grabbing information from in order to fill the area that it's replacing. Okay, so I'm happy with that. So I'm going to click apply, and that will keep me in the workspace. Then on the left side, on the original, we can scoot over to the right and I can use the lasso tool here in order to select this iceberg in the upper right. Here, we see the preview. It's done a great job with all these default settings. So I can click apply and then we can continue to delete other areas or fill other areas. But for now, I'm going to output these to a new layer, click okay, and we can see in our layers panel, there is the one ice and the other ice, and then our original retouching from the last video. Okay, let's merge these three layers down. I'm going to just hold down the shift key and select all of them, and then I'll use the layer menu and we can merge these because they're not overlapping and I want them all on the retouch layer. So I'll also just rename that "retouch". Then let's move to the snail image here. I want to remove this area here that looks like it's been a little bit distorted, so I'm just going to quickly select it with the lasso tool. And again, I'll select the edit menu and then Content-Aware fill. Now on the left side, we see the original, but on the right side we can see it's not doing a very good job of fixing or filling that area. So let's try changing the color adaptation to high. That helps a little bit, but let's also set the rotation adaptation to full to enable it to really rotate the pixels. Now it's giving me this little warning. I'm going to click don't show again, and then hit okay. So it's done a better job, but let's also let it scale the pixels, and now we can see that it's doing quite a good job. We could see if maybe just setting the rotation to full. Well, that looks good too. So let's go ahead and apply that. Now I want to compare that with a new technology, and that's generative fill. So I'm going to hide this layer for a moment and return to the background. So generative fill uses artificial intelligence to completely fabricate new pixels, and you do need internet connectivity in order to use the generative fill technologies. So here I can use the contextual task bar to quickly access generative fill, but you can also find it under the edit menu, right under the Content-Aware fill. When I select it, it will ask me if I want to enter in a prompt. In this case, I really want the rest of the photograph to be the prompt. So I'll just click generate. Now it's going to go up to the cloud, and it's going to generate three variations to fill in this area. And it's doing that, like I said, using artificial intelligence and machine learning. So we have these three variations and we can either click on them in the properties panel or you can toggle through them in the contextual task bar. So I think I like this one the best. If none of them were working, I could just click generate again and Photoshop would generate three additional variations. If I want to keep the file size low, then I can delete the variations that I'm not going to use. So those are the ones that don't have the blue highlight. And we can see in the layers panel, here's my generative fill layer. If I toggle that off and toggle on the background copy, you can see we get very similar results, but I actually think the generative fill is doing a little bit better of a job. So let's move over to the river jpeg. And here we can see I've got some leaves that were moving. This was an HDR image, so three exposures, and then I merged them together in Camera Raw. Let's zoom in to 100% and then pan over to this lower right area. If I were to use Content-Aware fill here, I'd have to be a little bit careful. In fact, I would probably go in with my clone stamp tool and make a more definite distinction between the river and the leaves. For example, I would option click maybe up here and then just fill in this area to help Content-Aware fill know that these are rocks and this is the tree that I want to remove, but I'll undo that for now because generative fill is going to do a much better job than Content-Aware. So I'm going to tap the L key that's going to gimme the lasso tool, and let's just select this area. In fact, I'm going to select into the rock area here, and then come down here and back up. And then we can hold down the shift key and look at, there's these leaves here. I don't really want them, and I don't really want this green or that green 'cause that's all distracting. Then I'll just use the contextual task bar, choose generative fill. I don't need to enter in a prompt and I'll just choose generate. It's nice that generate fill gives you these little tips each time you use it, and it's going to generate three different versions for us. So here we go. Here's the first one. Looks good. The second one and the third one. So if we zoom out, command 0, I actually think the third one looks best. I don't like the rock going all the way to the edge there. Yeah, I like this one. So I could throw away the other two. And then we could save this file as a PSD file or as a TIF file in order to keep the generative fill layer reeditable and separate from the background. So as you can see, both Content-Aware fill and generative fill are excellent tools for removing distracting elements in complex areas of an image.

Contents