Chapter 6 Data Processing: Software and hardware costs and options
Software costs We chose open-source options where we could, but there are still significant software costs. Table 6.1 outlines cost and explores options.
| Software | Purpose | Cost | Options - notes |
|---|---|---|---|
| DJI Terra Pro | Raw LiDAR processing | $6,000 plus $500 yearly | None, DJI LiDAR can only be processed in DJI Terra, and the yearly updates have included significant improvements |
| Agisoft Metashape Pro | Photogrammetry processing | $4,700 | Agisoft is industry standard. Pix4D is often compared, but more for online processing. Open Drone map (ODM) is a free open-source alternative, but has not been tested by this team. |
| LasTools | LiDAR processing | $3,100 | Excellent software that runs quickly, but most of the functions we used can be written in R. |
| RedToolBox | PPK and geotagging | $300 | PPK can be free in RTKLiB. Recently, the free Emlid Studio V1.8 update can PPK and geotag. |
Data management: Storage needs
Each project will have very different storage needs, but this will be a rough guide to help plan for the quantity of storage required and can be scaled according to project goals.
One flight per sensor:
L1 - 2Ha site produces ~510 pictures and approximately 8.5 GB of raw data total.
P1 - 2Ha flight area produces ~500 photos and 12 GB of data.
MS_P - 2Ha flight area produces around 12,000 photos equaling 42 GB.
H20T - 2Ha flight area produces aproximately 1500 RGB and thermal images equaling ~20GB.
We are keeping all intermediate products while we are developing processes. Processing more than doubles the raw data. Consider 250GB a conservative estimate for total data storage needs for the one day on site.
A copy of the folder structure we use is available on GitHub
Computer Hardware:
The computer we are using has the following specifications
Processor Intel(R) Xeon(R) Silver 4214R CPU @ 2.40GHz 2.39 GHz (2 processors)
Installed RAM 256 GB
System type 64-bit operating system, x64-based processor
GPU NVIDIA RTX A4000Raw Data Collection
2TB SSD for processing, we have found the processing is consistently faster and less prone to error if we process on the SSD rather than on the storage drives.
For storage we use an 8bay QNAP RAID6 with 12 TB hard drives.
We use 18Tb HDs for backups.