Pix4DMapper: Processing UAS Data

Introduction:

For this weeks lab our task is to create a orthomosaic image using Pix4Dmapper, this is a program used by geographers to create accurate 3D imaged maps. This makes the data very easy to read and very easy to analyze. The task at hand is to use this program and the UAS data provided to create a map of the study area and calculate different aspects of that study area.

Using Pix4D:

The most important thing to do when using Pix4D is to make sure the images are overlapping, when I say overlapping I don't mean a small amount of overlap, rather the higher the overlap the more accurate of a 3D model may be created. The reason why overlaps are so important when using this specific program is because it makes the 3D image more accurate and sharper than normal 3D images through the use of aerial triangulation(AT). The main goal of aerial triangulation is to determine the position and orientation of the images collected using UAS, this allows the AT to pierce through multiple images that are overlapping and create a legible map of the study area. In some cases aerial photographs may not capture as much information as needed, this is when the terrain is covered with either snow or sand. With these terrain types the overlap rates are much higher, with both snow and sand, this is because of how reflective these substances are and show minimal visual content, this is why there needs to be a certain amount of frontal and side overlap. The minimum levels of overlap for sand and snow are 85% frontal and 70% side overlap, these are minimal requirements meaning that these are the lowest values they can achieve for the images to be captured correctly. There is also another feature in which the data can be verified by the program, this is called rapid check. Rapid check is where all of the data is processes very quickly, the disadvantage to this method is that the results have low accuracy since all of the data was processes at a quick rate rather than being thorough. There are also a few more positives to using this program, for one Pix4D can process multiple flights at once, but only if certain requirements are met, the main one is that the coordinate system for all the images are the same. Another positive is that Oblique images may be processes through the use of this program, but only if the overlaps are of good quality and GCPs are present. GCPs are important but will not be used during this project, but they are a good option because when used they provide the results with more accurate findings. All of the results are presented in the quality report, this report is where all of the data is analyzed and all the figures are created.

Results:

The data that the class was provided with was UAS imagery data collected by Professor Josheph Hupy, the study area is a sand mine that is located in the south region of Eau Claire. To start off the  project all of the images were imported to Pix4DMapper, this is where the area of interest was chosen. The AOI was selected directly over the sand mine, the next step is to process the images and wait for the quality report to be generated. The creation of the quality report took about 15 minutes and the results were quite interesting. Below are figures 1 - 3from the quality report representing the information put into the report as well as maps showing the study area in different ways.

Figure 1: The quality report from the overlapping images being processed and analyzed

Figure 2: Represents both the Orthomosaic and DSM processed from the overlapping images

Figure 3: Represents the where in the study area there were strong overlapping between the images and weak overlapping

The areas in green represent the areas where there were multiple overlaps between the images making the results for that area very accurate and easy to read. While the yellow and red areas represent areas that contain poor overlap between images. Most of the areas that contain poor overlap are on the edges of the map, but the main point taken from the figure 3 is that as long the study area is showing strong overlap between the images than there will be positive results.

Conclusion:

Overall this was a very enjoyable and informative lab, learning and using a new program to create and analyze 3D images is expanding my knowledge on how to create legible and accurate maps. Pix4D is overall is a very user friendly program, making is easy to use for first timers. As well as one of the higher forms of technology for creating high quality results. This course has taught me a lot throughout the semester and I feel that ending on learning how to use a new program such as this is the perfect way to complete the course.

GPS Topography Survey

Introduction:

For this weeks activity the class was asked to take different points around a study area located on the UW - Eau Claire campus. Each point was taken using a survey grade GPS device, this allowed the class to get its points to within 5 meters of the real life location. Using this device it will make it much easier to collect data in the field. For each data point taken the elevation data was taken as well using the GPS device. Using this elevation data the class will be creating a continuous surface of the study area where a map can be created using ArcMap. The most important part of creating a surface area map is to take enough data points with a difference in relief points so than the elevation in the land is noticeable on the map. This will allow the study area to be correctly represented, the sampling method is important as well. In this case the class used a random sampling method, this was most important because this would allow the study area to be best represented in terms of the difference in elevation. The GPS device was portable and took points on real world locations using a tripod that needed to be level in order to correctly take the point.


Study Area:

Below is Figure 1 which is the study area that was captured by using google maps so it doesn't correctly represent the elevation of the area.

Figure 1: The Area where the elevation points were collected by the class

Location: Area in front of Centennial hall at the University of Wisconsin - Eau Claire campus. In the middle where the statue is located.


Methods:

The materials that were used in the survey was a Tripod stand that needed to be level, a Topcon Tesla Field Controller, and a Topcon Hiper SR Positioning unit. Below are figures 2 and 3 representing the materials used in the survey.

Figure 2: Topcon Hiper SR Positioning Unit on a level Tripod used in survey for collecting the elevation data points

Figure 3: Topcoln Tesla Field Controller used to take points while using the other device to take points

These materials used a bluetooth handheld device which was able to download the data collected onto a computer. Once all of the data was collected, Professor Hupy processed the data and shared it into a temporary folder that was in the format of a text file. From there the class needed to transfer the data on the text file to an excel file so than it can be used in Arc map to create maps that represent the data taken from the study area. The final step was to create five different maps each with a different interpolation. The five different types of interpolation are IDW, Spline, Natural Neighbor, Kriging, and TIN tools.

Results/Discussion:

Below are figures 4 through 8 representing each of the interpolation tools used in this survey.

Figure 4: IDW Interpolation of the elevation data points taken from the study area

Figure 5: Kriging Interpolation of the elevation data points taken from the study area

Figure 6: Natural Neighbor Interpolation of the elevation data points taken from the study area  

Figure 7: Spline Interpolation of the elevation data points taken from the study area

Figure 8: TIN Raster created from the elevation data points taken from the study area

After looking at all of the different interpolation tools that can be used on the elevation data taken from the study area, I believe that the Natural Neighbor interpolation method is the most accurate out of all five. It does the best at representing the differences in the elevation level in the study area, other methods did a good job on this as well but not as effective or accurate as the Natural Neighbor. 

Conclusion:

I believe that this survey was an overall success, the task at hand was to collect data points from a study area selected by Professor Hupy using a survey grade GPS device. Once data was collected it was time to transfer the data in order to create maps. The overall process was simple and easy to understand, the GPS device made everything with collecting the data easier than it should have been. But is was a great learning experience and gave the class a view at what is capable using a GPS device especially one that is so accurate to real life locations.