andrematte commited on
Commit
6fe953b
·
verified ·
1 Parent(s): 7732720

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +37 -3
README.md CHANGED
@@ -25,9 +25,25 @@ size_categories:
25
  This dataset contains a series of multispectral image slices captured at the embankment dams and dikes
26
  of the Belo Monte Hydroelectric Complex, located in the state of Pará, northern Brazil. Each image is
27
  paired with its respective NDRE vegetation index values, binary segmentation mask and multiclass
28
- segmentation mask
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
29
 
30
- ---
31
  ## Dataset Structure
32
 
33
  The dataset files are organized as following:
@@ -38,10 +54,28 @@ The dataset files are organized as following:
38
  ├── 📁 index_ndre # NDRE index values
39
  ├── 📁 labels_binary # Binary Segmentation Masks
40
  ├── 📁 labels_multiclass # Multiclass Segmentation Masks
41
- └── 📁 README.md
42
  ```
43
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
  ---
 
 
 
 
45
  ## Citation Information
46
 
47
  If you use this dataset in your work, please cite:
 
25
  This dataset contains a series of multispectral image slices captured at the embankment dams and dikes
26
  of the Belo Monte Hydroelectric Complex, located in the state of Pará, northern Brazil. Each image is
27
  paired with its respective NDRE vegetation index values, binary segmentation mask and multiclass
28
+ segmentation mask.
29
+
30
+ The multispectral images were captured by the Micasense RedEdge-P multispectral sensor embedded in a
31
+ DJI M210 V2 UAV. Radiometric calibration was performed for all images based on the known reflectance
32
+ values of a calibration panel. All images were used to process a Digital Ortophoto Map, which was then
33
+ sliced into the 256x256x5 image patches that are contained in this dataset.
34
+
35
+ Each image file is composed of six channels:
36
+
37
+ 1. Red band reflectance
38
+ 2. Green band reflectance
39
+ 3. Blue band reflectance
40
+ 4. Red Edge band reflectance
41
+ 5. Near-infrared band reflectance
42
+ 6. Binary cutline
43
+
44
+ The vegetation index (NDRE) values were calculated based on the spectral bands and the segmentation masks
45
+ were manually annotated using the CVAT software.
46
 
 
47
  ## Dataset Structure
48
 
49
  The dataset files are organized as following:
 
54
  ├── 📁 index_ndre # NDRE index values
55
  ├── 📁 labels_binary # Binary Segmentation Masks
56
  ├── 📁 labels_multiclass # Multiclass Segmentation Masks
57
+ └── 📝 README.md
58
  ```
59
 
60
+ ## Segmentation Classes
61
+
62
+ The image annotations are formatted for both binary and multi-class segmentation.
63
+
64
+ **Binary Segmentation Classes:**
65
+ - Slope
66
+ - Not-Slope
67
+
68
+ **Multi-class Segmentation Classes:**
69
+ - Slope
70
+ - Drainage Channels
71
+ - Stairways
72
+ - Background
73
+
74
  ---
75
+ ## License Information
76
+
77
+ This dataset is licensed under the [Creative Commons Attribution Non Commercial Share Alike 4.0 International](https://spdx.org/licenses/CC-BY-NC-SA-4.0) license terms.
78
+
79
  ## Citation Information
80
 
81
  If you use this dataset in your work, please cite: