A color image enhancement algorithm based on quaternion representation of vector rotation

Detail enhancement of color images is required in many applications. Unsharp masking (UM) is the most classical tool for detail enhancement. Many generalizing UM approaches have been proposed, for example, the rational UM technique, the cubic unsharp technique, the adaptive UM technique and so on. For color images, these algorithms have three steps: (a) Implement the color2gray step; (b) design an extracting method of high frequency information (HFI) based on the luminance component (LC); (c) complete the enhancing process utilizing the HFI. However, using only the HFI of the LC may lose the HFI of the chrominance component (CC). This paper proposes a quaternion based detail enhancement algorithm to extract details of the color image using both of the luminance and CCs. The proposed algorithm is designed to address three tasks: (1) designing an extraction method of the color high frequency information (CHFI) based on quaternion description of the 3D vector rotation; (2) performing an effective fusion strategy of the CHFI and the gray high frequency information (GHFI); (3) designing a quaternion based measure method of the local dynamic range, based on which the enhancement coefficients of the proposed algorithm can be determined. The performance of the proposed algorithm compares favorably with many other similar enhancement algorithms. The eight parameters can be adjusted to control the sharpness to produce the desired results, which makes the proposed algorithm practically useful.