IndexError: list index out of range in class ImageSet

[Seo-4d696b75]

Hi. I'm very new to EasyIDP, so I have no idea this error is a bug or fault of my usage.

Now I am following the tutorial in wiki: Basic Usage, and demo data is been used. In section 1.2.2 Transformer, I tried calling geo2raw.get_img_coords_dict but error happened. My source and error output is;

import sys
sys.path.insert(0, './EasyIDP')

from easyric.objects import Pix4D
from easyric.io import shp, geotiff, plot
from easyric.calculate import geo2raw, geo2tiff

import pyproj
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

# root dir of demote data is under the current dir
project_path = "lotus_170525"
p4d = Pix4D(
    project_path=project_path,
    raw_img_path=f"{project_path}/photos",
    project_name=project_path,
    param_folder=f"{project_path}/params"
)

plots = p4d.read_shp3d(shp_path=f"{project_path}/roi/plots.shp", get_z_by='all', shp_proj=pyproj.CRS.from_epsg(4326))
img_dict = geo2raw.get_img_coords_dict(p4d, plots['N1W1'] - p4d.offset.np, method='pmat')

Traceback (most recent call last):
  File ".\test.py", line 26, in <module>
    img_dict = geo2raw.get_img_coords_dict(p4d, plots['N1W1'] - p4d.offset.np, method='pmat')
  File "./EasyIDP\easyric\calculate\geo2raw.py", line 171, in get_img_coords_dict
    projected_coords = pmatrix_calc(param, points, im.name, distort_correct)
  File "./EasyIDP\easyric\calculate\geo2raw.py", line 75, in pmatrix_calc
    xyz = (xyz1_prime).dot(param.img[image_name].pmat.T)
  File "./EasyIDP\easyric\objects.py", line 261, in __getitem__
    return self.img[self.names.index(key)]
IndexError: list index out of range

Taking a look into class ImageSet, length of 2 lists: self.names, self.img are mismatched.

>>> len(p4d.img.names) 
266
>>> len(p4d.img.img) 
5

For now, I want to run my code and know the features of EasyIDP, I edited class ImageSet in easyric/objects.py as follow;

    def __getitem__(self, key):
        if isinstance(key, int):  # index by photo name
            return self.img[key]
        elif isinstance(key, str):  # index by photo order
-           return self.img[self.names.index(key)]
+           return next(filter(lambda i: i.name == key, self.img), None)
        elif isinstance(key, slice):
            return self.img[key]
        else:
            print(key)
            return None

Thank you for your answer.

[HowcanoeWang]

Dear seo,

Thanks a lot for using this package and ask your questions.

You don't need to change any source code, the problem is caused by: the pix4d project used 266 raw images to generate the dom, however, the dome data only has 5 images in it.

>>> len(p4d.img.names) 
266
>>> len(p4d.img.img) 
5

The simplest way to fix this problem, is downloading full raw images:

https://github.com/HowcanoeWang/EasyIDP.paper/releases

https://github.com/UTokyo-FieldPhenomics-Lab/EasyIDP/releases/tag/v0.1-alpha

Only lotus_170525/photos folder is needed, you don't have to download full 12GB data.

PS: In the easyidp v2.0, you can easily access previous data by #31 :

import easyidp as idp
lotus = idp.data.Lotus()
# photos
lotus.photos
# projects
lotus.pix4d.project

And the API has been significantly changed, please refer to the documentation for more details.

[UQDannySmith]

Hi @HowcanoeWang,
I am having the same issue, but it seems like the link above doesn't work. Do you have another link I could use?

[HowcanoeWang]

@UQDannySmith Please check the updated link. The lotus_170525.zip in the Assets should work