Resource-Optimized Delivery of Web Images to Small-Screen Devices

In this paper, a framework for adaptive delivery of web images to small-screen devices has been presented. The proposed system provides content-level scalability by unifying image analysis, compression, optimized content adaptation, and user interaction. We have proposed a way to establish the rate-distortion information for each representation of the objects. A feature space distance is introduced as a new distortion measure of image approximations. Additionally, new algorithms for color reduction and down-sampling have been developed, minimizing the distortion on the feature space.