A novel frequency domain approach to automated pseudo-colouring of images

Pseudo colouring is widely used to colour monochrome images or to artificially colour an existing colour image to highlight specific features in that image or to enhance visual quality. A novel automated pseudo-colouring technique is proposed in this paper. The proposed method comprises of two steps; reference image selection from the database and colour transfer using histogram interpolation. The first step relies on Content Based Image Retrieval (CBIR) using wavelet transform. The performance of our method is compared with other automated pseudo-colouring approaches through visual inspection based on MOS (Mean Opinion Score). The simulation results prove that the proposed method has many merits compared to existing techniques.