kg-ylrust/notes.txt at main · dcher95/kg-ylrust · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
DATA

- The data has different resolutions and spectral bands.

Spatial Resolution
- 4 bands at 10 m (B2, B3, B4, B8)
- 6 bands at 20 m (B5, B6, B7, B8A, B11, B12)
- 3 bands at 60 m (B1, B9)

Band	Resolution	Central Wavelength	Description
B1	60 m	443 nm	Ultra Blue (Coastal and Aerosol)
B2	10 m	490 nm	Blue
B3	10 m	560 nm	Green
B4	10 m	665 nm	Red
B5	20 m	705 nm	Visible and Near Infrared (VNIR)
B6	20 m	740 nm	Visible and Near Infrared (VNIR)
B7	20 m	783 nm	Visible and Near Infrared (VNIR)
B8	10 m	842 nm	Visible and Near Infrared (VNIR)
B8a	20 m	865 nm	Visible and Near Infrared (VNIR)
B9	60 m	940 nm	Short Wave Infrared (SWIR)
B10	60 m	1375 nm	Short Wave Infrared (SWIR)
B11	20 m	1610 nm	Short Wave Infrared (SWIR)
B12	20 m	2190 nm	Short Wave Infrared (SWIR)

EPSG:32630

Pixel Values: [0, 65535]

All data is in grayscale.

Different folders in training refer to specific plots, and different timeframes of photos taken.

*******
DATA SPLIT

- ~2000 timed photos in training/validation
- ~500 timed photos in holdout

*******

MODELING

- Which bands do I want to use?
    - 12 bands are given. 3 are typically used in most ImageNet pretrained models.

    If I choose to only use select bands:
        - Use RGB bands of most granular 10m.
            Band 4 (Red)
            Band 3 (Green)
            Band 2 (Blue)

        - Create false-color image for vegetation analysis:
            Band 8 (NIR)
            Band 4 (Red)
            Band 3 (Green)

        - Enhanced vegetation index:
            Band 8 (NIR)
            Band 4 (Red)
            Band 2 (Blue)

    Some pre-trained models using more bands are available in torchgeo:
        - They have already used SSL, and are available for 13 bands.
            Could use as a baseline, and adjust head for fine-tuning.


- How do I take into account the temporal effect?
    - Validation data doesn't have time identifiers, so probably not of consequence. Accounting for time wouldn't help because we don't have that information during evaluation.
    - Integrating time element would be good skills to learn.