From the course: PFTrack Essential Training

Auto match - PFTrack Tutorial

From the course: PFTrack Essential Training

Start my 1-month free trial Buy for my team

Auto match

“

- [Man] In this chapter we're going to take a look at all the different options we have for tracking objects within our scene. Let's go down to tracking tab and the first one we'll take a look at is the auto match. Let's select the auto match and bring it up to our tracking tree. So I just want to take a moment to thank the folks at Wolf Hollow for letting me film at the Wolf Sanctuary in Ipswich, Massachusetts. Really cool organization you can check them out at wolfhollowipswich.org and I'll put a link below so you can check it out. Really cool organization and a beautiful place if your ever in the area. The way that the auto match works is that it selects features on every single frame and then connects those features together to create tracks. This is slightly different from the auto track. The auto track will select features on a certain frame and then try to track those through. So a slightly different algorithm, a slightly different approach to auto tracking. So let's take a look at some of the options within the auto match. The first thing we can do is turn on the preview matching. And when we turn that on you can see all the features that the auto match is going to try to track. We have some options in the points tab. We have the current group which is the camera. We have our in and out points. Below that we have our target number and that's the amount of features that we're trying to keep on each frame. Below that we have the minimal length for the tracker. So right now, the setting is five. Any trackers that only last for four frames or lower will not be counted in the auto match. Then we have our matching options. If we set our settings a certain way that we like and we keep using we can set a pre-set for that at the top. Below that we have the different channels that the auto match will use. If one of the channels is really noisy you might want to turn that off to get smoother tracks. I know for example the blue channel a lot of times can be really noisy and we can turn that off and just track on the red and green. That can give us better results for the tracking. We have an option for image proxy if we are working with larger format like four K or above we can use a proxy to help speed up the process. Then we have our feature scale which we have a couple of options here. We have small, medium and large. You can see what happens when we change those. It will jump on to different features. I'm going to keep this at small. We have our feature strength. So that's the percentage of how strong the edges are. By default it's 50% which is usually pretty good. Then we have the matching area and that's similar to the tracking window. That's the size of pixels that it will look for the feature in the following frames. And when you press on the E button to edit the matching area so if we had a quicker moving shot you might want to make that larger. Then we have our motion prediction. If we had meta data for the camera, we could use that meta data or we can use the image or both. I'm going to clip it on use image and with those settings set. Let's click on the auto match and see what we get. Okay, so we've finished our first auto match. You can see at the bottom it gives you the average length of the trackers. So you can see here that 94.4% of the trackers are 50 frames or more. Now that we've already done the process we can actually change the length depending on different settings. So as I move this up you'll see that it will adjust. So that's an overview of how to use the auto match. Next we're going to take a look at using the auto track.

Contents