Joint 3-D Image Quality Assessment Metric by Using Image View and Depth Information Over the Networking in IoT

It is accepted that image compression and transmission are essential in the image processing system and that errors occurred may introduce the degradation on the quality of the received image over the wireless transmission for intelligent Internet of things (IoT). Image quality assessment metric (QAM) is of fundamental significance to various image processing systems, and the goal of studying the QAM is to design an algorithm that can automatically evaluate the quality of the received image at the terminal display equipment in a perceptible way under the ubiquitous network circumstances. In this paper, focusing on the image compression and transmission, a joint full-reference (FR) QAM ( JQAM) for evaluating the quality of a 3-D image is proposed based on the state-of-the-art physiological and psychological properties of the human visual system (HVS) for image transmission under the wireless networks. The major technical contribution of this paper is that the binocular perception (depth perception) and local image properties are taken into consideration. First, the luminance masking property and local image content information are calculated to establish image QAM (IQAM) to improve the abilities of current quality evaluation algorithms. Meanwhile, the information abstracted from the depth map is also utilized as side information (SI) to evaluate the depth map on quality assessment. Finally, IQAM and SI are combined to construct the proposed JQAM. Experimental results show that the proposed metric could achieve better correlation with the subjective quality scores compared with the relevant existing IQAMs, and it could be used to evaluate the quality of the 3-D image signal of the intelligent equipment for multimedia communication systems in IoT.