camera/vision for geo-location & geo-identification john s. zelek intelligent human machine...
TRANSCRIPT
![Page 1: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/1.jpg)
Camera/Vision for Geo-Location & Geo-Identification
John S. Zelek
Intelligent Human Machine Interface LabDept. of Systems Design Engineering
University of Waterloo
![Page 2: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/2.jpg)
Why can’t we use GPS everywhere?
Urban canyons
Indoor navigation
1. Introduction - 2/20
![Page 3: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/3.jpg)
What we are trying to do
CameraInertial
Altimeter, Compass+/- GPS =
Accuracy +Location +
Maps +1. Introduction – 3/20
![Page 4: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/4.jpg)
Applications
1. Introduction – 4/20
![Page 5: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/5.jpg)
SLAM
Given:Dead-reck.Ext. sensorWaypoints
Not Known:MapGPS
2. SLAM – 5/20
![Page 6: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/6.jpg)
Trees as landmarks
for triangulati
on
2. SLAM - 6/20
![Page 7: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/7.jpg)
Daniel AsmarSlide 7
Differentiating different trees
2. SLAM – 7/20
![Page 8: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/8.jpg)
2. SLAM – 8/20
![Page 9: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/9.jpg)
Object Category
Recognition
3. Object Detection & Recognition – 9/20
![Page 10: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/10.jpg)
Classes of Objects vs. Instances
2 instances of an individual object(space shuttle)
2 instances of an object face class
2 instances of an
object motorcycle
class3. Object Detection & Recognition – 10/20
![Page 11: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/11.jpg)
Visual vs. Functional classes
There is a wide variation in the
appearance of objects that are categorized
by function. We focus only on
categories related by some
visual consistency only!
3. Object Detection & Recognition – 11/20
![Page 12: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/12.jpg)
Challenges
changes of viewpoint
transformation (translation, rotation, scaling, affine), out-of-plane (foreshortening)
illumination differences
background clutter
occlusion
intra-class variation
3. Object Detection & Recognition – 12/20
![Page 13: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/13.jpg)
Ours
Others
Repeatability of our detector appears to be better!
3. Object Detection & Recognition – 13/20
![Page 14: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/14.jpg)
Object Graphs
3. Object Detection & Recognition – 14/20
![Page 15: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/15.jpg)
3. Object Detection & Recognition – 15/20
![Page 16: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/16.jpg)
3. Object Detection & Recognition – 16/20
![Page 17: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/17.jpg)
4. Structure from Stereo – 17/20
Structure from stereo
![Page 18: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/18.jpg)
Structure from motion4. Structure From Motion – 18/20
![Page 19: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/19.jpg)
5. Context Recognition – 19/20
![Page 20: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/20.jpg)
6. Closing – 20/20
![Page 21: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/21.jpg)
Extra. Features for Recognition & Structure – 21/20
![Page 22: Camera/Vision for Geo-Location & Geo-Identification John S. Zelek Intelligent Human Machine Interface Lab Dept. of Systems Design Engineering University](https://reader030.vdocuments.us/reader030/viewer/2022032607/56649ecb5503460f94bd98ee/html5/thumbnails/22.jpg)
Extra. Features for Recognition & Structure – 22/20