Human computer interactions for converting color images to gray

A classical image processing problem is converting color images to gray. This paper presents a Virtual Luminance Axis (VLA) to provide users a friendly interface to control the conversion process. The user selects several interesting color regions by drawing a few brush strokes, and then tunes the maximum and minimum of the corresponding band for each kind of color on the VLA. Our algorithm derives global mapping of the color space onto the VLA as close as possible. In this way, the user can “tune” the resulting image to produce the ideal effects he desires. Quantization and interpolating techniques are employed to accelerate the computation. The experiments show that the final gray is transferred as user defines.