With the advent of modern digital technology, users can capture an image and reproduce it between different media, such as to display it on LCD or CRT monitor, print it on desktop printer, or send it to a printing press. The challenge is then to maintain the accuracy of image colors during this reproduction. This has led to the development of color management systems. Using these systems, the color reproduction across media is accomplished using device ICC profiles that describe each device's color characterization data in a standardized format in terms of a device independent color space (profile connection space or PCS). ICC display profiles use a matrix transformation or a multidimensional lookup table (LUT) to map the PCS to the device colorant space. The matrix transform may be obtained by linear regression. The LUT, however, is usually constructed based on an estimated characterization device model (using nonlinear regression for interpolation functions fit to a set of measurement data) to speed the transformation performance. Due to the significant role that monitors play in color management systems, their characterization method needs to be accurate and reproducible. This paper evaluates existing display characterization methods for LCD monitors and uses the evaluation results to develop a new enhanced display characterization method that smooths the display device gamut and reduces measurement noise. A C++ program code is constructed to build a new well-behaved (continuous, differentiable, with continuous derivatives, and invertible) display profile, using the resulting values from the new characterization method. The AE error is computed to evaluate the accuracy of the characterization.
展开▼
