obsidian/Main
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

vault backup: 2025-03-26 16:31:49

Browse Source
This commit is contained in:
Obsidian-MBPM4
2025-03-26 16:31:50 +01:00
parent ad26715fb2
commit a2e42353d2
2 changed files with 7 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
3 Knowledge/Computer Vision - Depth Perception.md
View File

@@ -8,7 +8,7 @@ tags:
# Computer Vision - Depth Perception
## Stereo Vision
### Disparity Map
[This blog article](https://www.baeldung.com/cs/disparity-map-stereo-vision) explains nicely what disparity maps are.
[This blog article](https://www.baeldung.com/cs/disparity-map-stereo-vision) explains nicely what disparity maps are. And [this blog article](https://johnwlambert.github.io/stereo/) goes more into the math and physics of it.
- The disparity is the apparent motion of objects between a pair of stereo images[^1].
- The depth is inversely proportional to the disparity. If we know the arrangement of the cameras, then the disparity map can be converted into a depth map using triangulation.
- When disparity is near zero (far away) then small differences produce large depth differences. When disparity is large, small disparity differences do not change the depth significantly. Hence, stereo vision systems have high depth resolution only for objects relatively near the camera.

7
99 Work/0 OneSec/OneSecNotes/10 Projects/Requirements/Requirements Gathering.md
View File

@@ -9,4 +9,9 @@ tags:
- Ensure that cameras are well calibrated
- for stereo vision: vibration, relative motion of the cameras, when the wings flap, etc. See [here](https://youtu.be/GpU1Vx-b3VA?t=187).
- for the AI model
- #osd/question do we account for camera calibration in the dataset generation?
- #osd/question do we account for camera calibration in the dataset generation?
- parameter system: the drone needs to be fully configurable through a parameter system.
- For smooth delivery, we must be able to pull the payload bay away from under the payload in the last moments of the ejection procedure. Therefore, the payload axis motor/servo must be able to accelerate downwards with at least 1g  (9.81 m/s2) acceleration at the outer edge of the payload bay. Also the max speed has to be such that we can pull the payload edge down by ~4 centimeters in around ~50ms. Probably it's more helpful to focus on this max speed figure than the acceleration. From a standing start, the actuator needs to be able to pull the payload edge down by 4cm in 50ms
- payload system must be able to withstand strong external torque, from a human operator who is forcefully trying to open or close the payload bay by hand
- the payload axis actuator (motor, position sensor etc) has to be waterproof or waterresistant
- the payload axis actuator must have enough torque to close the payload bay under the most extreme payload mass condition. Which is roughly 1.5kg of payload mass, 10cm off axis, meaning 15kgcm of **dynamic** torque on the axis