The recent advancement of 3D sensing technologies has led to immense collection of voluminous and complex point clouds. Evidently, we are witnessing multiple developments related to assessing the voluminous point cloud data. The no-reference quality assessment is challenging but is widely accepted as a suitable quality estimator as it does not require the pristine stimuli. The no-reference quality metrics when applied directly to the point cloud fails to consider features that are close to the human visual system. Incidentally, visual quality is always driven by how scenes are perceived by the observer. Hence, there is a need to devise a novel quality estimation model which does not require the references, and exploits the perceptual features from the distorted point clouds. In this work, we propose a projection-based no-reference point cloud quality estimation model (PB-PCQA). The applied projection approach considers all possible viewing angles for a given distorted point cloud. Further, the model also extracted feature maps which are close to the human visual system from each 2D projection of the point cloud. In addition to this, a set of statistical parameters are extracted from each of the feature maps for estimating the distribution of the feature values. The proposed model is evaluated on benchmark SJTU-PCQA dataset and is compared with recent state-of-the-art metrics.