Gray level reproduction in dark areas is one of the major image quality issues of PDP(Plasma Display Panels) that
represents gray levels with pulse count modulation. Error diffusion and dithering have been frequently utilized for
smooth tone reproduction in dark areas. However, they may result in side effects such as the visibility of minor pixels,
flickers and/or motion artifacts. This paper presents a framework to evaluate the performance of gray level reproduction
algorithms for PDPs. Various factors affecting the quality of gray level reproduction in dark areas are introduced.
Spatiotemporal luminance variations are proposed to measure the performance of the gray level reproduction.
Experimental results indicate that the proposed method faithfully represents the performance of gray level reproduction
capability of PDPs.
The CRT has the nonlinear light intensity response to the input digital values. However, light intensity curve of the PDP is close to linear within the normal operating range. Thus, in order to generate images on the PDP being equivalent to those on the CRT, the modification on the input digital values should be made. This process is often called as inverse gamma correction. When the inverse gamma correction is applied, the number of displayable gray levels in dark areas is considerably reduced. It would result in false contours that reduce the quality of displayed images. Error diffusion can be applied to solve this problem. However, the luminance variation between two consecutive integer values is much greater than 'just noticeable difference' of human vision system. It makes minor pixels appear as isolated dots. In this paper, an error diffusion algorithm is proposed to reduce the luminance variation. Also, three algorithms called in this paper as 'Fraction doubling', 'Minor pixel separation' and 'Within-frame minor pixel allocation' are proposed.