Abstract:Interactive color transfer is an important image editing technique, which changes the color pattern of a given source image following users' intention and is more flexible than fully automatic methods in control of color modification. This paper presents a new interactive transfer approach which consists of two stages. Firstly, users specify the color change style by drawing some strokes in the source image and their reference strokes in one or multiple reference images. The source strokes should respectively obey the color style of their reference strokes after color transferring. This is actually a color transfer problem too, which is then casted as a quadratic optimization forcing the color statistic attributes of the source strokes to follow those of their reference strokes while preserving their gradient map unchanged. The source strokes are separately treated and therefore they can bear different color styles after processing. The solution of the optimization is very efficient due to the fact that the specified strokes with the same color style usually contains a small number of pixels. Secondly, the new colors of pixels in source strokes are propagated to the whole image automatically. Instead of solving an optimization equation, we renew the color attributes of each pixel in the source image using a scattered interpolation function which is expressed as a weighted combination of the new colors of pixels in those specified strokes, where the weight is inverse to the multiplication of the appearance distance and position distance from the current pixel to the pixel in the source strokes. Experimental results show that the proposed approach can not only preserve the texture detail and reflect user's intention well but also achieve realtime performance even for large images.

Interactive Image Recolorization Based on Chrominance Blending

Color Transfer Between Images with Auto Selecting Swatches

Improved method of image color transfer.

Language-based colorization of scene sketches

Colorization For Chinese Ink Painting

Fast Colorization Using Edge And Gradient Constrains

Color Transfer Using Scattered Point Interpolation

Optimization Based Grayscale Image Colorization

Control Color: Multimodal Diffusion-based Interactive Image Colorization

Self-similarity-based image colorization

User Tailored Colorization Using Automatic Scribbles and Hierarchical Features

Image Recoloring Based on Fast and Flexible Palette Extraction

Hand-drawn Grayscale Image Colorful Colorization Based on Natural Image

A learning-based approach for automatic image and video colorization

Patch-driven colorization

Automatic Image-Based Pencil Sketch Rendering

ColorizeDiffusion: Adjustable Sketch Colorization with Reference Image and Text

SmartPainter:Colorizing Grayscale Images Using Diffusion Colors

Deep Palette-Based Color Decomposition For Image Recoloring With Aesthetic Suggestion

Intrinsic Colorization

Palette-Based Image Recoloring Using Color Decomposition Optimization