Multitasking Image Editing System

Image editing technique improves greatly these years, and more and more researchers all over the world begins to think up with new ideas and solutions to various problems in this area. Color editing, size editing and content editing are mainly discussed in this paper. New framework for image cloning is presented, and algorithms for color adjusting and image resizing are improved. And a GUI system is implemented for image editing. A number of experiments are carried out to prove the feasibility of our algorithms.In the area of color adjusting, color harmonization algorithms is improved in this paper. And our system can define arbitrary color model according to practical use. And hue value is adjusted according to this model, in order that the color of the whole image is changed. The fixed image models are discarded, and better color performance is achieved. A GUI system of arbitrary color editing is implemented in this paper, besides a colorization system is made in order to colorize a grayscale image.In image resizing area, content aware image resizing algorithm is improved in this paper, and a user interactive system is implemented. Critical object in the image can be preserved, while the aspect ratio changes. And this leads to a result without distortion. A concept of seam is introduced in this algorithm, and image size can be changed by delete or insert these seams, with the content of the image reserved. The image enlarging algorithm is improved in this paper and the "ill stretch" problem is solved. This algorithm can be applied in content emphasizing and object removal etc. And the result can be guided by the user, and the system can run in real time as well.In image cloning area, we present a method called arbitrary cloning in which image matting technique and Poisson image editing are applied to accomplish the image seamless cloning. First, the object boundary is extracted through image matting technique. And Poisson equation is solved to achieve natural result for seamless cloning. Compared with state-of-art techniques, the paper gains advantages that great performance can be reached in bad conditions that there is complicated color variation in the foreground and background image; theαmatte can be guided by the user input. The algorithm is very flexible in application. For current used cloning technique, the algorithm guarantees better output, higher speed and diverse results.