Object tracking developed with Kalman filter & A*

The object tracking was developed with Kalman filter. It works good on simple situation tracking and still working to fix the problem like crossing, drastically change.

A* algorithm, part of collision avoidance method, was implemented by former program.

I've spent almost 1 month on it. It works! YEAHHHHHHHHHH

Fusion, recording, sensor information

1. Fusion

It has some information of sensors shown below,

SV	ESR
Rects	Detected points
Detected average points

2. Date retrieving

It can record _each sensor_ or all of them at the same time and so does _replay_.

3. OpenCV cuda

It's implemented on OpenCV3 beta on my computerXD.

4. Stereo vision

Information of X, Y is calculated followed stereo vision theory.

5. Road elimination

Prototype is implement, but not robust for all situations. I think v-disparity may be helpful.

bias of disparity, video recording class, GUI updated

Unknown bias of disparity is fixed
video recording class
GUI is updated

Topview prototype

Fusion_v1.0

Topview information of sensors