A novel approach to multi-scale blending based on saliency mapping for multimedia image compositing applications

In multimedia applications, compositing is an inevitable step for creating special visual effects from entertainment to education. It is often desirable to merge two regions or an object from one image over background of another image to get a completely new image. However, compositing multiple images with different textures and colors results in the creation of visual artifacts along the boundaries. A multi scale seamless image compositing method is proposed to minimize these artifacts in order to generate visually appealing and natural looking composites. Unlike existing compositing techniques, the proposed method automatically detects the visually salient objects from the source image for merging it over a destination image background. Novel facets of the algorithm are: (1) use of a novel salient object detection method for automatic mask generation, (2) formulation of a novel 2-D exponentially weighted Savitzky-Golay (2-D EWS-G) filter, which is an extension of 2-D S-G filter, for multi-scale pyramid based blending with edge preserving blurring property. Experiments performed over different datasets are compared with the state-of-the-art methods to justify the effectiveness of the proposed methodology. Despite smoothing the seams, our algorithm retains the sharpness of the image features, effectively reduces noise, and is robust to change in window size of the proposed filter. Thus, the proposed algorithm exhibits a considerable improvement in image compositing and hence it could be quite useful for multimedia applications. (C) 2016 Elsevier Ltd. All rights reserved.