Object Detection and Tracking using Sequential Images

We extract information of motion from sequential images, and process it.
Our goal is to detect and track moving objects by using optical flow, which is
velocity field in an image arising from an motion between camera and object.
We basically use "Gradient Method" based on spatio-temporal filtering.
Theoretically, optical flow can be extracted by using two different spatial filters,
but we use four spatial filters to reduce error. Details are here in case of
two filters,
four filters.
Researches
-
Realtime Moving Object Tracking based on Optical Flow
-
Realtime Person Tracking based on Optical Flow
-
Realtime Multiple Objects Tracking based on Optical Flow
-
Person Tracking Using Opticalflows and Homogeneous Regions
-
Alien Detecton based on Optical Flow
-
Object Tracking based on Optical flow and Depth
-
Object Contour Tracking based on Optical flow and Edges
Realtime Image Processing
We use the realtime image processors
"ISHTAR",
"ISHTAR-II", which use several DSPs(Digital Signal Porccessro), to process
images from camera in real time. We extract optical flow and use it to detect and
track objects in real time.
Furthermore, we control the active camera head
"PANTHER-VZ"
to pursuit an target object using results of tracking. This enable to keep
the target in the view of camera.
Group members
Papers
Ryuzo Okada, Yoshiaki Shirai and Jun Miura
"Object Tracking based on Optical Flow and Depth", Proc. of IEEE/SICE/RSJ
Int. Conf. on MFI, pp.565-571, 1996
Ryuzo Okada, Yoshiaki Shirai and Jun Miura
"Object Tracking based on Optical Flow and Depth", Proc. of RWC
pp.432-439, 1997
Ryuzo Okada, Shinya Yamamoto, Yasushi Mae Yoshiaki Shirai
"Realtime Person Tracking System"
Yasushi Mae, Yoshiaki Shirai, Jun Miura, and Yoshinori Kuno
Object Tracking in Cluttered Background Based on Optical Flow and Edges
Proc. 13th Int. Conf. on Pattern Recognition, Vol.1, pp.196-200, (1996).
Yasushi Mae, Shinya Yamamoto, Yoshiaki Shirai, and Jun Miura
Optical Flow Based Realtime Object Tracking by Active Vision System
Proc. 2nd Japan-France Congress on Mechatronics, vol.2, pp.545-548, (1994).
Back to SHIRAI Lab's Home Page
okada@cv.mech.eng.osaka-u.ac.jp