Hierarchical Objective Quality Assessment for CS Video in WMSN

Abstract

Compressive sensing (CS) is a sub-Nyquist sampling way while still enabling exact reconstruction, which is applicable to WMSN. In this paper, based on the characteristic of CS video in WMSN, we proposed a hierarchical objective CS video quality assessment (HOCSVQA) approach to get CS video quality index (CSVQI) from three levels, measurement level, stream level, and packet level, respectively. This approach cannot only keep the convenience and real-time characteristic of objective video assessment, but also reflect the QoE to a certain extent due to the coefficients regressed from subjective video assessment experiments. A set of experiments on subjective CS video quality assessment and another set of verification experiments are designed and settled. The CS video quality index, CSVQI, assessed by the model we proposed maintained a high correlation with data from verification experiments under statistical correlation measure.