I couldn’t find anywhere in the general handbook of instructions where it says who should join the bishop on the stand. I know who I’d want on the stand with me, but I don’t know if it’s allowed.
I’m looking for recommendations on algorithms / libraries for deduplicating moving object detections across two unsynchronized RTSP camera streams viewing the same scene from different angles. The goal is to determine when detections from camera A and camera B correspond to the same physical object.
Constraints:
- Cameras are not hardware synchronized so frames may be offset in time
- objects can be from 30 to 150 pixels with squarish aspect ratios, are moving in frame, with varying speed along mostly consistent trajectories
- object detector recall is ~65%
- I have intrinsic + extrinsic calibration between cameras with roughly pixel level reprojection accuracy
- Intra-camera tracking is working reasonably well. My harder problem is cross-camera association: matching sparse trajectories between two vertically stacked, calibrated cameras viewing the same moving objects. Motion is mostly transverse in the image, detections are intermittent, and some objects may only have 1–2 detections in one view.
What suggestions do you have?
I bought a “Lemon Elberta seedling” from my local nursery. I wanted a true Elberta, but that was the closest they had in stock. I don’t see a graft in the root so I suspect it did come from a pit. Has anybody had any experience with stone fruit from seedlings? How can the nursery know what the fruit will be if it was grown from seed (with unknown male lineage)?
I bought a “Lemon Elberta seedling” from my local nursery. I wanted a true Elberta, but that was the closest they had in stock. Has anybody had any experience with this type of seedling? How can the nursery know what the fruit will be if it was grown from seed (a pit)?
I know the Camino has been mostly secularized (one Reddit discussion estimated participant motivation was 85% secular) I’m curious what percentage of pilgrims are openly religious, e.g. grace at meals, nightly individual prayers, scripture reading, church attendance, alburgue led group prayers, etc.
If you are religious, did you feel comfortable expressing your religion among other pilgrims on the Camino?
Do some routes draw more religious pilgrims than others?
What are some unique hostels on the Camino Frances? Communal meals? Atmosphere? Evening activities? Sense of community?
I have some sub stream video that has image frames that are 352 x 240 and looks perpendicularly at a road. I have been unable to find a pre-existing model that can detect license plates in small images and oblique angles. I don’t need to read the plates, just detect them. However, every model I’ve tried has failed miserably. I’ve looked at Roboflow, huggging face and GitHub. Alternatively, maybe somebody knows of a license plate dataset with non straight-on samples that I can sub sample and train on.
Thanks for the help!