Local tone mapping using K-means algorithm and automatic gamma setting

Tone mapping (TM) algorithms reproduce the high dynamic range (HDR) images on low dynamic range (LDR) display devices such as monitors or printers. In this paper, we propose a local TM algorithm, in which the input HDR is segmented using K-means algorithm and a display gamma parameter is set automatically for each segmented region. The proposed TM algorithm computes the luminance of an input that is the radiance map generated from a set of LDR images with varying exposure settings. Then, according to the bilateral filtered luminance, an image is divided into several regions using a K-means algorithm. The display gamma values are set automatically according to the mean value of each region. Then, the tone of HDR is reproduced by a linear TM method with adaptive gamma value. We generate the tone-mapped image by the proposed local TM. Computer simulation with real LDR images shows the effectiveness of the proposed local TM algorithm in terms of the visual quality as well as the local contrast. It can be used for contrast and color enhancement in various display and acquisition devices.