Investigations on logarithmic gamma funtions for the storage and post-processing of high dynamic range image data

OData support
Supervisor:
Mócsai Tamás
Department of Networked Systems and Services

Traditional LDR (Low Dynamic Range) cameras can only capture the fraction of the world, what a human is able to see. Thus there is a growing demand to capture HDR (High Dynamic Range) content. During professional shootings the aim is to capture the most of the information possible for the postproduction, where the creator needs the maximum freedom and quality to create the best impression at the result.

However HDR content means a huge amount of data and needs enormous computing capacity at the post-processes. Plus nowadays systems only support 8-10 bit depth contents. So logarithmic gamma functions are made to store the HDR content in lower bit depths, thus at smaller space and to solve the compatibility problems with the extant systems. These characteristics are good alternatives for the traditional gamma correction characteristic to quant HDR content at 8-10 bits.

The topic of my thesis is the investigation on the logarithmic gamma functions for the sore and post-process HDR content. My task is to investigate on how big quality loss is caused on the content at the storage and post-process by the usage of these functions compared to raw processed content.

The thesis summarize the theoretical knowledge needed for the investigation, staring with the HDR content and content production. Then I present the traditional gamma characteristics, and the logarithmic gamma functions. Afterwards I demonstrate the post-procession steps what is used for the investigation. I introduce the objective quality measurements what I used to make the results and the test environment I implemented in MATLAB. Finally I analyse the results made by the implementation.

Downloads

Please sign in to download the files of this thesis.