Research In High Dynamic Range Image Encoding Method And Image Base Lighting

Because of its excellent performance in visual effect, the high dynamic range (HDR) image has revolutionized the field of computer graphics and related areas. With the develop of computer graphics, HDR technique will become more and more popular. The problem is that most of the display devices available nowadays are still for low dynamic range (LDR) content only. As a result, find a way to achieve the compatibility between LDR content and HDR content is necessary. At this point, the first task in this paper is that propose a method which allows typical drivers to read and display a tone mapped image while the HDR information is carried in "sub-band marker". On the other hand, it can recover the HDR image with the help of inverse tone mapping operator (iTMO) plus "sub-band marker". The model is the backwards-compatible enhancement to HDR. When an application is not designed for HDR imaging, it displays only the tone mapped image, allowing the user to have access to part of the content. When an application is designed for HDR imaging, the model can recover the HDR image by invers tone map operator.A critical process of the compatible encoding model is designing iTMO algorithm,which is used to transform LDR content into HDR content. This paper propose an iTMO algorithm base on frequency separation with the goal of preserving edges and local contrast. The separation is achieved by using bilateral filter, then a linear expansion and brightness enhancement are performed for low frequency part. Finally, the expanded layer, and the high frequency part are recombined to form a new HDR image. The experimental result suggests that our algorithm is better than many other methods.Image-based lighting (IBL) technique helps people integrate synthetic object into real world environment. It is one of the most important applying of HDR technique, which use HDR image as lighting. However, the general IBL model is hard to achieve expectation result, since it rarely consider the shadows and reflect lights which is caused by synthetic objects. In this paper, the second task is to improve the general IBL technique by taking the interacted effect between the envirement and synthetic objects into account, this can be achieved by creating supporting plant for synthetic objects, then supporting plant will show shadows and reflect lights. Besides, the proposed method can modify the background image slightly for special purpose base on texture transplanting. The created texture is saved in database, when people wish to modify the background image they call texture from database and transfer it to the target surface. The improving IBL method can produce reality visual effect.