Add Support for DJI ILS corrections #52
Conversation
| @@ -22,4 +22,4 @@ dependencies: | |||
| # subpackage dependencies | |||
| # Sentera dependencies | |||
| - pip: | |||
| - git+https://github.com/SenteraLLC/py-image-metadata-parser.git@v1.18.1 | |||
| - git+https://github.com/SenteraLLC/py-image-metadata-parser.git@add-dji-ils | |||
There was a problem hiding this comment.
Will update after pushed to main
| if make == "DJI": | ||
| return imgparse.get_irradiance(row.image_path) | ||
| else: | ||
| return imgparse.get_ils(row.image_path)[0] |
There was a problem hiding this comment.
I need to compare the results from both of these options and maybe fallback on ILS values if the irradiance values are not there. The Irradiance values are much more consistent throughout the flight and from what I've read the irradiance values are based on sun angle and other info stored in the images from the DLS(Directional Light Sensor). There isn't much info that I've found that compares the 2 values and a lot of stuff I've found makes it seem like they are the same.
| @@ -65,7 +65,7 @@ def extract_panel_bounds(image): | |||
| corners, ids, rejected_img_points = cv.aruco.detectMarkers(image, dictionary) | |||
|
|
|||
| # if at least one marker detected | |||
| if ids is not None: | |||
| if ids is not None and (ids[0][0] == 23 or ids[0][0] == 63): | |||
There was a problem hiding this comment.
I was getting an error for aruco ID 157 detected and think I've seen customers with this issue, I looked up that ID and its super simple to idk maybe something in the Mavic 3 sensor writes that pattern and it is being picked up incorrectly. This will skip any images that have the wrong ID detected. The image I had was not even a calibration image.
Add Support for DJI ILS data
What?
Why?
PR Checklist
__version__and pyproject.tomlBreaking Changes