The bandwidth and storage restrictions of consumer devices and conventional delivery infrastructures on stereoscopic 3D video require efficient compression methods to save the bandwidth and preserve the perceptive quality at the same time. Compression efficiency is actually determined as the visual quality achieved for a certain amount of bitrate. In this paper, a perception aware coding scheme is proposed to improve the perceived quality of stereoscopic video. The proposed method exploits the sensitivity of human visual system to perceptual characteristics of texture video, to further remove the perceptual redundancy of stereoscopic image pair. The macro blocks within a frame are classified according to their overall brightness, for each of which an adaptive local quantization parameter is selected. The idea is that the darker macro blocks within an image can be compressed with higher ratio without sacrificing the visual quality, especially in the 3D video, where the viewers are more attracted to the pop-out regions and their motion in depth. The proposed scheme has been evaluated by subjective testing, confirming that no quality degradation is introduced by deploying the proposed method. In fact the perceived quality of the stereo content remains intact, while the objective measurements show a Bjontegaard delta bitrate reduction of up to 37.83% and with an average of 27.27%.