Photography is involved in modern communication more than ever and good communication value of photography has a deciding influence on successful communication.
In our research we determined communication value with the help of image quality assessment, that was based on subjective method (web application, crowdsourcing), objective method of eye tracking, mathematical objective measures (RMSE, PSNR, SSIM) and colour differences. Assessment was based on a testing image database that we developed for this purpose. Database includes 30 photographs with 28 different manipulations that are based on different visual quality parameters.
We determined that higher sharpness, lower contrast and lower saturation have little influence on image quality and on the other hand, higher contrast, change of lightness and lower sharpness have big influence on it. We calculated strong correlation between results gathered with subjective method and ones gathered with an eye tracking method, but the correlation with other objective methods is less significant.
In our research we also developed two new methods, one is for subjective quality assessment and the other for analysing data gathered with the method of eye tracking.
|