AT 309 Week 12: Pix4D Part 1

Introduction

Up to this point, AT 309 has taught us how to use GIS software and how to properly optimize data, in these software, so it can be analyzed. We now know how data is collected and analyzed, but there is an important step in between data collection and analysis that we have not learned about. The remainder of the semester is focused on this in between part: image processing. Data is not ready to be uploaded in to GIS software immediately after a field outing. GIS software need input in the form of point clouds, digital surface models (DSMs), and orthomosaics. These inputs are created through image processing which GIS software do not do. Pix4D is the software used in Purdue UAS for image processing. In week 12, we are being introduced to Pix4D and we do not do any image processing in this lab.

Help Documentation

While Pix4D is relatively easy to use, it is not a software you can use correctly without prior knowledge of sensors, image processing, and geospatial data. There are many options in Pix4Dmapper and it can get confusing to someone that is new to the software. For users that are having issues processing their data, there are several tutorials provided by the Pix4D team. 

Pix4Dmapper Software Manual

The software manual provides users with detailed explanations for every option available in Pix4Dmapper. The software manual can also be downloaded as a pdf for users that need offline access to the manual.

Pix4Dmapper Video Tutorials (YouTube)

There are also several video tutorials that show off the features of Pix4Dmapper.

Adding Data

Add Images Box

This is one of the first boxes that pops up when starting a new project. This is where images are added. Once images are added, Pix4D tells the users if there are enough images to do image processing (green checkmark at the top).

Geolocation Information (Image Properties)

Images may not always be geotagged directly in UAS missions. Geolocation data can come from cameras, GPS systems, and flight logs from certain software. In cases where images are not directly geotagged, a geolocation file needs to be added. Pix4D tells users whether their images are geotagged in the box shown above. Geolocation data can be added in this box as well. 

Output Coordinate Systems

Previous lectures in AT 309 taught the importance of coordinate systems. The image above shows the box where users can select output coordinate systems. Pix4D will autodetect the coordinate systems of the data provided but users may want to change the output coordinate systems. This is where that can be accomplished.

Processing Templates

The final box that pops up before image processing can begin is the "Processing Options Template". There are many options here. The main one used in the future image processing labs is "3D Maps". 3D Maps is used for mapping missions where the camera is pointing straight down.

Processing

Processing in Pix4Dmapper is completed through four phases: 1) Initial Processing, 2) Analyzing the Quality Report, 3) Point Cloud and Mesh, and 4) DSM, Orthomosaic, and Index. The quality report does not show up as one of the phases in Pix4Dmapper. Instead, the quality report is generated after each phase is completed and it gets updated along the way. The quality report details a lot of information about the image processing. It is very useful to use when identifying if data is being processed correctly. Additionally, users can know when their processing is complete because a pdf version of the quality report will appear in results folder for their project.

GCP/MTP Manager

Another menu, that is used in the final lab of the semester, is the GCP/MTP manager. Here users can add GCP data, edit GCPs, and mark them. GCPs are important because they provide accurate geolocation data. The GCP Manager needs to be used to calibrate this GCP data. 

Conclusion

It is important that users know how to use Pix4Dmapper before they jump in to a project. This lab demonstrated some of the features in Pix4Dmapper. It also showed us that there are plenty of resources available should we run in to a problem in the image processing labs. Pix4Dmapper is not that difficult to use and, after this lab, the class should be ready to begin image processing.