Tying GCPs to images

- [Instructor] Now that I've covered how to incorporate GCPs into your project, let's use them in Drone2Map towards the creation of a True Ortho. You should have a CSV of the ground control points in the exercise folder for this chapter. And if you haven't downloaded and unzipped it, you should do that now. We are going to tie the GCP points to the drone images in this project. And the process is a bit time consuming, but the results are simply spectacular, if you do a good job. If the image viewer is still open from the previous video, please close that now. Also, we're going to right-click on the flight point right here, since it's active, and we're going to choose to clear it. Now it isn't selected. Now, over on the main menu home tab, we want to find the control group. Here, I'm going to click the dropdown for control and then control again, and I want to choose import control to import the GCPs. Now I want to navigate to the GCP file. Here, I put it in my desktop. So, This PC, and Desktop, into our project folder, and there it is. Select it and click okay. Here, you will set the parameters for the GCPs. Now these points are in a different coordinate system than the one that's currently shown as WGS 84 UTM Zone 16 North. So we need to change this to GCS NAD 83 (2011). And we can do that by clicking on this globe icon right here, and then choosing geographic coordinate system. In this case, we want North America. We'll scroll down here to USA and territories, and NAD 83 (2011), and I'll click okay. Now, it says that the geographic coordinates will generate UTM products when processed, which is completely fine. We just want to make sure that the GCP points have the right projection. Now these other fields down here were populated by the CSV, latitude, longitude, elevation, and label, but we did not include accuracy horizontal and accuracy vertical, so let's set them now. I'm going to choose this dropdown here and choose H_Accuracy and then V_Accuracy. Okay. The results of importing of the GCPs reveal the locations as green plus icons. You can see them here in the layers view, under control. And remember, these are where I placed those GCP flags on the ground and captured their location. Now, if you revisit the same drone project areas frequently on your many projects, it really is best to get the GCPs of permanent structures instead of flags, since structures don't change, but flags could slightly shift when trying to place them in the same spot each time on revisit drone projects. Now the next step is to identify the GCP locations in all of the images. Yes, you heard it right, in all of the images. It is a bit of a tedious process, but the results are worth it. Let me show you how to do it, and then you can proceed to do the work with the video paused. Let's go ahead and uncheck the image centers and the flight lines in the content pane here, so that they won't interfere in the GCP tie process. On the right, you have the control manager pane open with the GCPs we brought in from the CSV. Let's select the first GCP row by clicking it. At the top of the control manager pane, click the first icon, which is the show image links editor. In this editor window, you will go through and identify the location of the first GCP in the image with the flag on the ground as your helper. If you move your image links editor window over a little, you can see the GCP 1 plus right here on this map. So in this first image here, of the DJI_0373 image, if you zoom in, you'll find the GCP flag right here. There it is. I'm going to move this over here so I can see the whole thing, and know which image I'm on. Now basically, what we want to do is we want to left-click in the very center of this flag. And now you can see on the left here, for that image, we have a GCP set. And we can confirm that here by looking at the control point type, GCP, ground control point. Okay, now let me show you the next one. I'm going to click the next image here in the image viewer. I will zoom down to the flag, and then I will click in the center of it. And we're going to repeat this process for all of the images here that have this ground control point flag. Okay, so it looks like none of the rest of the images have that GCP, just all of these up here. Did you find them all? So, when done setting the GCPs for flag 1 to the correct images, we can click okay right here in the lower right. Nice job. You've just tied a precise location to several of the images in our project. And now let's do that for each of the other five flags. Now in the control manager pane on the right, you have the GCP 1 with a green check. Here it is. So that tells us that that's complete. We're going to select this second one here, and open up the image links editor. Now, ground control point 2 is right here. Here's the flag, and that's me, and this is where I launched the drone from this project. So, here I'm going to zoom in and set the center. And now I'll repeat the process for the other images here in the image links viewer. Okay, it looks like we're done with the image links viewer for ground control point 2. I'll go ahead and close that, and now we see a green check for GCP 2. Okay, now we're going to do GCP 3, and then 4, and 5, and 6 in the exact same way. So, nice job. You've just linked real world coordinates into the images that were flown in this area. Now, if you don't know the coordinate information features in your image to the exact location, you can always use checkpoints instead of ground control points. In the control manager, you see each one of these types as a GCP, but let's imagine we don't have these flags in our project area. If there were other features on the ground, such as painted X's on the concrete or street hole covers or other easily identifiable things, we can do the exact same process we just did, but instead we use check points, CPs. The process is the same, but instead of setting them as the type GCP, it's set as check point in the image links window viewer. So here, instead of having ground control points, you would set it as check point. Now, check points can be used to authenticate the overall absolute accuracy of a product, and Esri says that check points are collected similarly to control points, but are not incorporated into the block adjustment process. Instead, control points are used to independently assess the accuracy of the model after the processing is completed. They also say that the bundle block adjustment process attempts to fit a model to the location of the control, and as a result, the accuracy of the product is highest at the control points. So on this project we're doing together, we're just going to use the GCPs for our processing.