Understanding 20190318 Orb Slam2 Test 2
Exploring 20190318 Orb Slam2 Test 2 reveals several interesting facts. 20190318-ORB-SLAM2-test-2
Key Takeaways about 20190318 Orb Slam2 Test 2
- Monocular RGB only (no depth information) used. Notice tracking is lost quickly. Not a robust mapping solution. See the following ...
- The map was created using RGBD (+depth), yet relocalization works reasonably well with only RGB. See the map created in the ...
- SLAM is the computational problem of constructing or updating a map of an unknown environment while simultaneously keeping ...
- Frames captured with an iPhone in RGB and processed with a machine learned algorithm to infer depth information.
- Example of mapping a room and saving the map. After capturing, the frames were processed with a machine learned model to ...
Detailed Analysis of 20190318 Orb Slam2 Test 2
Simulate PAN TILT STILL ZOOM. ORB SLAM #2 TEST 20190318-ORB-SLAM2-test-1
Tuning depthMapFactor ...
Stay tuned for more updates related to 20190318 Orb Slam2 Test 2.