Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
|
@@ -10,4 +10,23 @@ pinned: false
|
|
| 10 |
license: mit
|
| 11 |
---
|
| 12 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
| 10 |
license: mit
|
| 11 |
---
|
| 12 |
|
| 13 |
+
This project is result of submission to the ETH AI Center GainForest Hackathon for XPrize.
|
| 14 |
+
|
| 15 |
+
The project repository contains Deepforest and DETECT-TREE2 models, which are image segmentation models.
|
| 16 |
+
|
| 17 |
+
Initially, the App asks user to upload an .jpb Panoramic image, it is automatically segmented and the output is a excel file with co-ordinates.
|
| 18 |
+
|
| 19 |
+
The segmented trees from the panoramic images are compared with identified species and results are generated on this.
|
| 20 |
+
|
| 21 |
+
The identification/classification is done using a VGG-16 model which is pretrained on 1 million images.
|
| 22 |
+
|
| 23 |
+
Testing of this model is done by a Dataset created for this project: https://segments.ai/r/drone-panoramic/.
|
| 24 |
+
|
| 25 |
+
This dataset has species annotated in the panoramic images. The results of the classification from the VGG or any other model can be tested with this dataset.
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|