29 January 2007 Threshold value for acceptable video quality using signal-to-noise ratio
Author Affiliations +
Noise decreases video quality considerably, particularly in dark environments. In a video clip, noise can be seen as an unwanted spatial or temporal variation in pixel values. The object of the study was to find a threshold value for signal-to-noise ratio (SNR) in which the video quality is perceived to be good enough. Different illumination levels for video shooting were studied using both subjective and objective (SNR measurements) methodologies. Five camcorders were selected to cover different sensor technologies, recording formats and price categories. The test material for the subjective test was recorded in an environment simulator, where it was possible to adjust lighting levels. Double staircase test was used as the subjective test method. The test videos for objective measurements were recorded using an ISO 15739 based environment. There was a correlation found between objective and subjective measurements, between measured SNR and perceived quality. Good enough video quality was reached between SNR values of 15.3 dB and 17.2 dB. With 3CCD and super HAD-CCD technologies, video quality was brighter, less noisy, and the SNR was better in low light conditions compared to the quality with conventional CCDs.
© (2007) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Mikko Vaahteranoksa, Tero Vuori, "Threshold value for acceptable video quality using signal-to-noise ratio", Proc. SPIE 6494, Image Quality and System Performance IV, 649405 (29 January 2007); doi: 10.1117/12.702995; https://doi.org/10.1117/12.702995

Back to Top