I would like to know how does Kinect work and how it is different from Nintendo Wii motion system?
Kinect is based on the hardware, and possibly software, of PrimeSense. They use a flashing IR light-source and a filtered camera to measure time-in-flight and compute a depth map from that. Presumably, objects that are near-field and/or moving are detected as people, while static and far-away stuff is treated as background.
In contrast, the Wiimote has an IR camera and accelerometers built into it. The sensor bar that you put on/near your TV contains a pair of IR LEDs. The hardware in the Wiimote tracks those LEDs and determines which way it's pointing based on their position.
Kinect is very different from the Nintendo Wii.
The Nintendo Wii requires you to be holding WiiMotes and only cares where the WiiMotes are and what they are doing.
Instead, Kinect tracks your body in 3 dimensions and tries to do 1:1 replication of your movements. So if you jump, Kinect will see it and make your character jump on the screen.
Kinect also has built in headset-free chat, so you can talk to the game/other players freely.
Kinect does not use time of flight. It uses an IR projector and an IR camera. You can even see the results of the IR projector using a camcorder with night vision. The most thorough description I have seen is here which includes footage.
Given that the Kinect is capturing a scene and then doing image processing on the result there is the possibility that it will make more mistakes when calculating your position and movements than the Wii which relies on the remote. While I'm sure that the developers at Microsoft have done all they can, it would pay you to make sure that you follow the instructions very carefully when siting the Kinect bar etc.
The Wii remote has an infrared camera which needs to see the 2 IR LEDs on the sensor bar to work out it's position and orientation. It sends this together with the accelerometer data to the console. The remotes have a definite range - beyond which they either can't detect the sensor bar or the apparent distance between the lights is too small for it to calculate it's position.