readme.md

This document will introduce you how to do the batch processing step by step. It totally has 5 steps:

1. Test project to find out whether config works
2. Batch loading images and marker detection
3. Batch align photos
4. Change bounding box (bbox) of processing area
5. Run reconstruction and export results

Link: your/data/root/path

.
├── 00_rgb_raw
│   ├── broccoli_tanashi_5_20211021_P4RTK_15m
│   │   ├── DJI_0224.JPG
│   │   ├── ...
│   │   └── DJI_0226.JPG
│   ├── broccoli_tanashi_5_20211025_P4RTK_15m
│   ├── ...
│   └── broccoli_tanashi_5_20220412_P4RTK_15m
├── 01_metashape_projects
│   ├── bbox.pkl
│   ├── broccoli.files
│   ├── broccoli.psx
│   └── outputs
└── 02_GIS
    └── gcp.csv (also can be any name you want)

Step 1: Test project

The key features of this step are as follows:

• Test whether config.json and scripts works on your data.
• Find out the correct ID between GPS coordinates and coded targets.
• Define the bounding box of plot.

Please follow these steps:

1. Open an empty metashape project, remove the Chunk 1

   and save the project on your disk

2. Edit the config.json in 10_agisoft_batch_tools/config.json

   {
       "img_path": "your/data/root/path/00_rgb_raw",
       "test_folder": ["subfolder_a"],
   
       "skip_folder": ["subfolder_a", "subfolder_b", "subfolder_c"],
       "ref_csv": "your/data/root/path/02_GIS/gcp.csv",
   
       "bbox_pkl_path": "your\\data\\root\\path\\01_metashape_projects\\bbox.pkl"
   }

   To accelerate the test step (no need to load all flights), only recommend use one or two subfolder (flight) as test. The script will only load test_folder into the test project.

   The skip_folder contains some bad flights and should never be used.

   Please note, only / or \\ are acceptable for the path string, the Windows default paste 'E:\aaaa\bbbb' doesn't work and can cause error.

3. Create the gcp.csv in your data root path 02_GIS

   By default, using the RTK GPS device to measure ground control points in field, may get the table like this:

   

   the FID is often not corresponding to the coded target id.

   Please copy and paste the XYZ data to gcp.csv, the columns are (the order is very important!):

   id	longitude	latitude	altitude
   0	139.5405953	35.73646447	96.881
   1	139.5410378	35.73957293	96.584
   ...	...	...	...

   PS: No need to add the header line into csv file:

     0, 139.5405953, 35.73646447, 96.881
     1, 139.5410378, 35.73957293, 96.584
     ... 
   

   PPS: The id here can be any distinguisable id, either FID or 点名 or any other non duplicated string. Then we will edit it again later.

4. run 10_agisoft_batch_tools/00_create_tested_project.py in metashape

   

   it auto adds metashape flights as chunk

   

   Also it will display the UAV positions (dots) and ground control points positions (flags), however those auto-detected "target_x" are not shown.

   

   No worry about it, we will display it later. (01 will solve this double detecting problem)

5. Edit the flight height of all the chunks

   

   Important: You need to manual setting for ALL chunks, no API for automatically edit this currently.

6. Then load the 02_align_photos_all_chunks.xml script by:

   

   If you meet the error can not load the script, please either update your metashape version (Recommended), or edit the xml file:

   <?xml version="1.0" encoding="UTF-8"?>
     <batchjobs version="1.8.0">  // edit here to match
   ...

   Then you the target_x are also appears in the software:

   

   Please edit the gcp.csv according to the previous relationship, eg: taget43<-> 3, then change 3 to 43.

   

7. Finally, change the plot bounding box to a proper size

   

   Then run the 03_change_bbox.py script like step 1.4 mentioned. You will save the adjusted bounding box to *.pkl file for later use.

   

8. Then the test project can be removed

Step 2: Batch apply to all

1. Open a new metashape, delete default Chunk 1 and save project as Step 1.1
2. Run 10_agisoft_batch_tools/01_add_images_to_projects.py as Step 1.4
3. Manually adjust the Capture distance for each chunks, as Step 1.5
4. Batch process 10_agisoft_batch_tools/02_align_photos_all_chunks.xml as Step 1.6
5. Manual check if all chunks are in good condition
6. Run 10_agisoft_batch_tools/03_change_bbox.py to batch modify all chunks' bbox
7. Batch export results by 10_agisoft_batch_tools/04_batch_full_outputs.xml, the results will save to your/data/root/path/01_metashape_projects/outputs

Step 3: Append flights

The previous pipeline can be run again when you have new flight data. But the already existed chunks (flights) will be ignored. Please DO NOT change images that already been exported and processed into the project.