There are several numbers famous proposed Block Matching Algorithm (BMA) in video coding technique and among it, the ARPS is a well known BMA technique that produce lower computational complexity and higher quality of the encoded video at the same time. In general, a video will has a lot of temporal redundancy among its neighborhood frames especially for a low motion video which make encoding a low motion video with smaller MB and bigger p size seemed impractical and vice versa. In this paper, the hybrid version of ARPS technique is used depending on its motion video type either low, medium, or high motion video. Basically this hybrid model works by setting the Macro Block (MB) and Search Range Size, p according to the motion type. Low motion video will be use higher size of MB and smaller size of p, medium motion has medium size of MB and p, and high or fast motion video will use smaller MB and bigger p size. The experimental result shows that by using the hybrid BMA technique, it can produce a better quality of the constructed frame and also it achieve less computational complexity at the same time.

No abstract

In this paper, a method to identify hand gesture trajectory in constrained environment is introduced. The method consists of three modules: collection of input images, skin segmentation and feature extraction. To reduce processing time, we compare the absolute difference between two consecutive frames then choose which frames have the highest value. YCbCr colour space is selected as the skin model because it behaves in such a way that the illumination component is concentrated in a single component (Y) while the blue and red chrominance component is in Cb and Cr. The hand gestures trajectory is to be recognized by using two methods: template matching and division by shape. Template matching required the removal of the head of the signer, leaving with just 2 hands only. For division of shape, the gesture are grouped into 5 classifications of hand postures that is vertical, horizontal, 45 0 above, 45 0 below and overlapping with hands. A total of 43 frames were selected manually for each hand posture and analyzed to obtain the variation of hand gesture feature such as width, heights, angle and distance. Our experimental results show up to 80% of accuracy in identifying the forms of the gesture trajectory. It shows that the feature extraction method proposed in this paper is appropriate for defining particular gesture trajectory.