lab 5 Pix4D Processing data

on
Introduction
  • What is Pix4D?
    • it is a computer software that can find different tie points or common areas in multiple pictures.
  • What products does it generate?
    • orthomosaics, digital surface models (DSM), and 3-dimensional models such as point clouds or meshes.

Part 2 methods 
  • What is the overlap needed for Pix4D to process imagery?
    • For this software it is recommended a 75% overlap in the picture direction and a 65% overlap in side to side.
  • What if the user is flying over sand/snow, or uniform fields?
    • Because the snow and sand tend to make images look more alike you need a 85% overlap in flight direction and at least 70% side overlap because many of the photos will look similar.
  • What is Rapid Check?
    • A rapid check make the tie points fast so instead of a higher quality image you get a lower resolution but faster made image.
  • Can Pix4D process multiple flights? What does the pilot need to maintain if so?
    • You can do multiple flights but the pilot needs to make sure he's using same altitueds and similar sensor to prevent errors there and more importantly make sure he get enough overlapping photos to tie the flights in.
  • Can Pix4D process oblique images? What type of data do you need if so?
    • Yes it can and for oblique images you need to add in ground control points(GCP's) and enough overlap.
  • What is the difference between a global and linear rolling shutter?
    • global takes the photo instantly 
    • linear rolling shutter takes the photo a little bit after it scans to capture an image.
  • Are GCPs necessary for Pix4D? When are they highly recommended?
    • GCP's are not necessary for Pix4D but a highly recommended when combining nadir images with ether aerial oblique and terrestrial images. Also recommended if capturing a tunnel system and you cant do multiple tracks
  • What is the quality report?
    • a quality report gives you all the information give from the processing in Pix4D. So it gives you information like how many image were used, the average number of tie points per image, previews of the results, details about the block adjustment, and other useful information
Part 3 methods
Figure 1
  • As seen in Figure 1 the first thing we did was add the data to Pix4D. Once we did that it gave us the points in which the the images were captured. At this point you don't have much useful information on the ortho.\

Figure 2
  • Figure 2 is when i did the initial processing. This is where it stiches together the ortho and chooses what images should be used. As you can see on the bottom of the page its shows the progress in the initialization phase.

Figure 3
  • Figure 3 is after the initial processing is done you get the quality report.  When the processing is completely done you will get it again but that one will be more complete.

figure 4
    Figure 4 is what the ortho looks like after the processing is done and while its in the point cloud form. you and use the triangles option to tie the points together and make a smother image as seen in the video.

    Part 3 Results and Discussion
    • In the end after all the processing we got our complete quality report. From it we can see that only one image was not used in the processing. The table is as seen in Figure 5. 

    Figure 5
    • After all was done we created a video as seen in Figure 6 in order to make the video we had to choose the overall amount of time we wanted it, the definition of the video. Then we when and choose how we wanted the fly by done by marking points in video selection and then editing them for the best video possible. you can see the route in which the camera takes and the angles it faces in figure 7
    Figure 6Figure 7

    • After we where done with the Pix4D we used the data we collected and uploaded it to ArcMap to make a map. We used the orthomosaic and the DSM to make the maps. At first I was going to make 2 different maps, for the DSM and for the ortho. which you can see my DSM map in figure 8. in the end i decided to use the DSM as an add on to the ortho. This is nice because you can look at the regular map the glance at the DSM to see the hights of the objects in question. The final map for this lab is Figure 9
                                             Figure 8                                                            Figure 9

    Part 4 Conclusion
    • Pix4D is very important for processing data of UAS systems because of all the capabilities and benefits that it has. With the right tools you can collect the proper data to be used and just like that you have the abiltiy to create accurate maps of a feature, digital surface models, and if you want to impress your client you can make 3D point clouds, Then using the animation tool create a video to showcase what all is possible with Pix4D. The entire ortho used 66 of the 67 images and took one hour and 7 minutes to do all the processing. And that is with a computer that has a large amount of processing power. This is where the problems occur in Pix4D because by the time your done processing the data it may be old and irreverent by that time. And that is where the biggest drawback occurs. This processing made the cpu's on our computers work so most regular computers may not even be able to handle the processing.    









    Comments

    Post a Comment