Full Processing vs Rapid / Low Resolution

Follow

Pix4Dmapper has Processing Templates with predefined options that can be used for processing. The default Templates (3D Maps, 3D Models, Ag Multispectral, Ag Modified Camera, Ag RGB, Thermal Camera) are designed to give the best outputs depending on the dataset. There are also the fast versions of these Templates that are designed to give faster, lower accuracy outputs (3D Maps - Rapid/Low Res, 3D Models - Rapid/Low Res, Ag Multispectral - Rapid/Low Res, Ag Modified Camera - Rapid/Low Res, Ag RGB - Rapid/Low Res, Thermal Camera, ThermoMap Camera). For more information about the Processing Options TemplatesProcessing Options Default Templates.

 

  Rapid / Low Resolution Full
Quality Reduces the resolution of the original images (depending on the dataset) and therefore lowers the accuracy and may lead to incomplete results. Fewer keypoints are extracted on each image and therefore the amount of matched points between the images is lower. Allows to process the original images with the best resolution (depending on the dataset) and therefore provides results with higher accuracy.
Speed It provides results significantly faster. It provides results significantly slower. 
Use It is recommend to use it: It is recommend to use it when back in the office to take detailed, accurate outputs.
 
Note: 
  • If the rapid / low resolution processing succeeds, it is safe to assume that the results of full processing will be of high quality.
  • If the rapid / low  resolution processing fails, it is an indication that this dataset requires more overlap (for more information about how to design the flight plan: Step 1. Before Starting a Project > 1. Designing the Image Acquisition Plan). If so, it is advised to collect more images either by flying again and combining the projects together, or by changing the flight plan to get more overlap.
  • It is possible to obtain results with the full processing even if the rapid / low resolution processing fails. However, the results in that case may be of lower quality, consist of less calibrated images or contain artifacts.
  • It is possible that the rapid / low resolution processing succeeds and the full processing fails. It is a rare case and usually happens with blurry or low textured images where few keypoints can be computed in full pixel resolution. It is advised to collect new sharp (not blurry) images. The full processing can succeed setting the Keypoints Image Scale to Half image size or Quarter image size. For more information: Menu Process > Processing Options... > 1. Initial Processing > General.
 
Important: When merging projects, it is recommended that the individual projects have been processed in Full mode with the same processing options.
 
Was this article helpful?
17 out of 17 found this helpful

Article feedback (for troubleshooting, post here instead)

6 comments

  • w DM

    Hi, will the full processing and fast/low resolution have any impact on the error of introducing GCP and CP?Looking forward to your reply. Thank you!

  • Avatar
    Momtanu (Pix4D)

    Hi,

    The RMS error for GCPs also depends on the reconstruction quality. Our software tries to adjust the reconstruction using the GCPs meaning that the goal is to minimize the RMS error. It is a local indicator of how well Pix4Dmapper fitted the model to the GCPs. If the reconstruction is not good, the RMS error will be high. The reconstruction will be different for full processing and rapid processing. Thus., the GCPs will also have different RMS error.

     

  • pascal TSEKPUI

    s'il vous plais quelles sont les indices utilisés pour déterminer la structure ou la texture du sol à partir des images multispectrales prises par drone? et quelles sont les formules? j'ai une caméra multispectral sequoia

  • Fernanda Bosmediano

    I have replied to your ticket. I will post this for other users. All the vegetation indices that you can generate using Pix4Dfields are listed on this page. There is the possibility to add a new index while using Pix4Dfields. However, for this, I would recommend you consult with your agronomist expertise. They will provide you more information about the best index for your soil.

    I have seen some studies where they use a vegetation index for this, but they use more information than an index. Here I leave you some articles that might help you:

    • In this study, they identified soil texture classed using an NDVI index, Digital elevation model. With a Sequoia camera, you have these bands: Green, Red, Red edge, NIR, Alpha. Therefore, you can generate the NDVI. 
    • Here, they show an "index of texture" based on determinations of moisture contents at the point of stickiness (P), and of sand contents (S) of soil samples. 
    • This study uses multispectral data and machine learning algorithms.                                                                                                                                      

    I would kindly ask our users to post on our Community  or create a ticket for support, we will gladly help you

    Edited by Fernanda Bosmediano
  • pascal TSEKPUI

    Bonjour 

    n'est-il pas possible d'utiliser les images collectées en 2D sous pix4D mapper?

  • Avatar
    Momtanu (Pix4D)

    pascal TSEKPUI It is. You can use 2D images to create a 3D model, point cloud, DSM, and orthomosaic in Pix4DMapper.

Article is closed for comments.