No-reference point cloud quality assessment based on subjective and objective scores

Recently, thanks to the increasing capability of 3D acquisition devices, point cloud has emerged as the most popular format for immersive media. In practice, a variety of distortions could be involved and affect human perception. Developing point cloud quality assessment can help to understand the distortions and carry out the quality optimization for distorted point clouds. Therefore in this work we are investigating the quality assessment of point clouds and what are the main factors effecting the quality of the point clouds. And presented as well the existing subjective and objective methods for evaluation of the quality of point clouds and applying these methods to evaluate the point clouds quality. After that, we made different experiments to apply the subjective and objective methods to estimate the quality of different data sets we have, And as well using the results of previous experiments made for farther investigations . However, applying these methods for estimating the quality is quite challenging approach as we will see later in this study. Therefore, we created various solutions to overcome the challenges that we faced during this research . At the end, we are investigating the using of the Neural Networks in the quality assessment of point clouds and how this can simplify the measurement of point clouds quality.