A method for emendation of attention trajectory in video content analysis is disclosed. A motion refinement module generates a refined motion vector for each macroblock of the plurality of macroblocks, based on the motion search motion vector. The first indoor heat exchanger contributes mainly to heating, and the second indoor heat exchanger contributes mainly to cooling. Then, an exterior skin mold is formed using the product prototype mounted on an alignment block.