Abstract:
High Dynamic Range (HDR) imaging aims to capture all the luminance values from real world scene in a single shot which is quite challenging. Display devices are not capable of reproducing all the luminance values from a natural image as the sensors in them are incapable of registering the range of brightness levels. The process of adapting the dynamic range of a real world scene or a photograph in a controlled manner to suit the lower dynamic range of display devices is called tone mapping. We use bilateral filtering approach and present a novel local tone mapping technique for high dynamic range (HDR) images taking texture and brightness as cues. Enhancing the details in an image to give a more pronounced look of the image is known as detail exaggeration. Conventional edge preserving filters such as the bilateral filter are a good choice but they introduce bright halos across the edges which look ugly. We propose an algorithm for content-aware detail exaggeration using local-Laplacian filtering which mitigate this shortcoming of the bilateral filter. While doing so, we make use of the saliency algorithm and image compositing techniques to achieve the detail exaggeration results. We also work on content-aware non-photorealistic rendering of images to achieve a desired depiction of the scene. We propose a novel content-aware framework in order to render an image for applications such as detail exaggeration, artificial blurring and image abstraction. In this work, we make use of the GrabCut algorithm and use Guided filtering for detail exaggeration application. We make use of real time image abstraction technique to give a cartoon like appearance to the desired part of the image. We conclude the thesis by presenting with a study on two tone suppression for audio watermarking where we explain the masking experiments required for corroborating the theory of two tone suppression phenomenon and its subtle use for watermarking technique. This is explained taking into consideration the MPEG-1 Psychoacoustic model-1 compression algorithm.