-
- Refactor implementation (minimize parameters, ...)
- Implement both seaming and carving as alternatives
- Stabilize stitching (preset output size)
- Remove artifacts (leftover texts, rope) with b…
-
UDIS real datasets doesn't seem to have validation datasets.
How did you validate the model and pick the best model when fine-tuning the homography network?
-
Four cameras were calibrated separately without joint calibration,
so getExtTranslation() is not useful.
Why not consider putting 4 cameras in the same world coordinate system to jointly calibrat…
-
## 🚀 Feature
Improved image batch plots during training
## Motivation
It's nice to see labeling performance as training progresses, but there are a few issues:
* Boxes are not coloured per-c…
-
There is an expected support for multispectral images in ODM as per [ODM's doc](https://docs.opendronemap.org/multispectral) and as from [PR-1057](https://github.com/OpenDroneMap/ODM/pull/1057)
To e…
-
Hi!
First of all, thanks for maintaining this great project.
I'd like to know if there is a way to input detections from multiple sensors to the tracker. My use case is that I have two sensors th…
-
Our documentation and tests include a variety of
examples and tests involving randomness.
Those depend on a randomness seed, and up to Sage 9.1
they always used the same one: 0. Thus every run
of `…
-
你好,我正在閱讀Learning edge-preserved image stitching from multi-scale deep homography,想請問ground truth offsets是怎麼來的?
![Capture](https://user-images.githubusercontent.com/67545227/192971100-8a49fa0b-a586-43…
-
Hello, I used viewselection.py to correctly generate viewpair.txt. Other configurations are the same as yours. I also used the Tensorflow1.13.1 version, and the cuda version is correct
However, the f…
-
Hi,
I have some uses cases where I need the map data beforehand for a specific scene. Here are two code examples:
```
def load_map(dataset: UnifiedDataset, scene: Scene) -> GeometricMap:
…