Automatic quality verification of the TV sets
Dusica Marijan, Vladimir Zlokolica, Nikola Dj. Teslic, Vukota Peković, Miodrag Temerinac · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
In this paper we propose a methodology for TV set verification, intended for detecting picture quality degradation and functional failures within a TV set. In the proposed approach we compare the TV picture captured from a TV set under investigation with the reference image for the corresponding TV set in order to assess the captured picture quality and therefore, assess the acceptability of TV set quality. The methodology framework comprises a logic block for designing the verification process flow, a block for TV set quality estimation (based on image quality assessment) and a block for generating the defect tracking database. The quality assessment algorithm is a full-reference intra-frame approach which aims at detecting various digital specific-TV-set picture degradations, coming from TV system hardware and software failures, and erroneous operational modes and settings in TV sets. The proposed algorithm is a block-based scheme which incorporates the mean square error and a local variance between the reference and the tested image. The artifact detection algorithm is shown to be highly robust against brightness and contrast changes in TV sets. The algorithm is evaluated by performance comparison with the other state-of-the-art image quality assessment metrics in terms of detecting TV picture degradations, such as illumination and contrast change, compression artifacts, picture misalignment, aliasing, blurring and other types of degradations that are due to defects within the TV set video chain.