Visual attention guided images abstraction using sparse coding

Given the increasing number of mobile platforms, a key technical challenge is how to provide efficient image processing application on resource-limited mobile devices. This paper proposes a novel technique for mobile image abstraction on mobile platform. In order to reduce the computation complexity, the image abstraction process is conducted on a cloud computing system. In this technique, captured images are analyzed to detect visual salient area, which is then provided for adaptive detail preservation of abstraction. To further reduce the image size for efficient transmission between mobile device and cloud computing system while maintaining the visual quality of abstracted image, the derived algorithm regards the salient regions as the dictionary in sparse representation, and selects the salient regions that minimize the residual output error iteratively, thus the resulting regions have a direct correspondence to the performance requirements of the given problem. Experimental results obtained using extensive datasets captured under uncontrolled conditions show the efficacy of the proposed system for conducting image abstraction on mobile platform using sparse representation.