You can easily know more about kinekt sensor from docs: https://msdn.microsoft.com/en-us/library/jj131028.aspx?f=255&MSPPError=-2147217396 (kinekt sensor exports depth measure in mm)

As regards "depth" it is the computed distance between the object and the kinekt sensor

Images are captured in 2D. Depth element gives more data about the other perspective (3D) such as thickness and orientation.

Kinect gives two videos include RGB and depth, with size 640*480. Depth frames or images have one channel and gray scale value from 0 to 255 . If the Object be closer to camera, in the image will be more brighter than another and its value is near to 255.