1. Process Image from stereo camera

2. Filter Skin, so we have only 3 point’s Head and two hands

3. Get Disparity (diferrence distance between image)

4. Get World Position (X, Y and Z)

5. Analyze motion

6. Transform from cartesian world coordinate to vector position so we have alpha, beta and tetha

7. That’s input for kohonen

Here is Abstract of  My D4 final projects :
People are the central element in the whole enterprise of multimedia and communication. The ability of computer program to find and follow’s people head, hands, and body is therefore an important visual problem for computer. Some games that needs people movement as interface will make that visual problem become very important.

In this final projects will use algorithm to let computer know where head, hand and body movement and position. That algorithm will implement on games with two cameras as interface. Images from each camera will process to get decision about movement type where at final will use it as a game controls.

Here is my final project demo video