Computer Vision Final Project - SURF Tracking - Ramp
Loading...
Loading...
1:48
Computer Vision Final Project - SURF Tracking - PETS2001 Camera 2by kurtglastetter3,030 views- 0:30
Object tracking algorithm merging SURF and LK Optical Flowby ehdvormf16,998 views
- 0:41
Tracking of 2D objects via SURF pointsby ArnaudRamey6,344 views
- 0:47
Eye gaze direction indicator v0.2 (CCA, blob tracking, OpenCV)by HTMOVIES60,694 views
- 0:21
OpenCV feature tracking using mobile GPU-based moboby kymhorsell2,093 views
- 1:04
FAST corners and SURF tracking on iPhone 3Gby carelvwyk3,856 views
- 0:52
OpenCV [SURF] Optical Flowby Pinuven2,263 views
- 0:45
Markless AR using SURF(SIFT) methodby dongwork20,991 views
- 0:35
Object Tracking On SURFby KimJunsaek1,918 views
- 0:52
gpu_surf_algorithm.mp4by jonathanclarkcom1,708 views
- 0:50
Real time tracking of a number plate using SURF descriptors on the GPUby RBFGFX4,849 views
- 0:39
Fast Object Tracking (Ball Tracking) with the OpenCV Libraryby elctronix56,040 views
- 0:34
Optical flow test: OpenFrameworks + OpenCVby perevalovds11,380 views
- 0:59
Computer Vision Final Project - SURF Tracking - PETS2000by kurtglastetter396 views
- 2:51
OpenCV_Surfby enddl221,498 views
- 0:44
UAT OpenCV SURF Point Trackingby BannisterRail550 views
- 5:51
Advanced object tracking algorithm and its application to auto mosaic programby ehdvormf2,606 views
- 1:57
Retrieving the center and contour of an object with OpenCVby Evol4ever3,218 views
- 1:22
Computer Vision Projectby davotop4,131 views
- 0:47
LED Tracking Using OpenCV and Frequency Measurementby jaggar28816 views
@AhoOo85: I used "Fast Scale Invariant Feature Detection and Matching on Programmable Graphics Hardware" by N. Cornelis and L. Van Gool
kurtglastetter 1 year ago
@AhoOo85, I did not implement GPU SURF.
kurtglastetter 1 year ago
Hi you got some nice speed going there. Did you implement the GPU-SURF ? If not , do you know where I can find it? I am doing my thesis project would be nice to try this out.
AhoOo85 1 year ago
@trianguloGS: Correct. Not using the SURF algorithm, that is, at least on current CPUs, which have relatively few cores. SURF is very processing intensive, but it is highly parallelizable, and particularly suitable for single-instruction, multiple-data (SIMD) processors, such as GPUs (or GPGPUS). However, there are other tracking algorithms that work quite well on CPUs (e.g. mean-shift tracking using probabilistic color, Lucas-Kanade tracking using Shi and Tomasi corner detection, etc.).
kurtglastetter 1 year ago
so with a CPU it can't do tracking in real time... ¿?
trianguloGS 1 year ago