The mean-shift algorithm is an efficient technique for object tracking [23]. Object tracking is achieved by applying it to the Bhattacharyya coefficient to find the object position. A bilateral filter [24] is used to overcome a number of problems, such as occlusion by an object with a similar color distribution or a background with a similar color distribution when an object is tracked using color information only. The bilateral filter consists of two support maps in the depth domain and the color domain. The use of the depth domain allows for easy segmentation of the user from a scene with other people and background objects. It is quite useful under changing illumination conditions to combine the advantages of color and depth data.