In the recent years, the improvement of digital imaging technology is significant. The capture resolution of consumer level digital cameras has grown significantly. This remains large space of improvement in the scope of color depth and dynamic range expansion....[ Read more ]

In the recent years, the improvement of digital imaging technology is significant. The capture resolution of consumer level digital cameras has grown significantly. This remains large space of improvement in the scope of color depth and dynamic range expansion.

In color depth expansion, the bit-depth of each color component in every pixels and expanded. In low color bit-depth images, the captured real scene information is quantized for storage. As some of the high precision details are lost, “contour effect” is introduced in gradient region creating false edges to the smooth area. Some general bit-depth expansion methods which perform direct mapping are available. However, these methods do not deal with the contouring effect.

In dynamic range expansion, the dynamic range of the capture image is expanded. Limited by the capture sensors and storage media, the dynamic range of captured image are restricted. In order to expand the dynamic range, every pixel components have to be captured in multiple exposure settings. The multiple captured values are combined to give a high precision, wide ranged pixel value. This is known as the High Dynamic Range (HDR) problem.

In this thesis, research works in color depth expansion and HDR are described. A color bit-depth expansion algorithm concealing contour edges is designed. This is followed by a dithering method which extending the bit-depth expansion algorithm. Finally, Spatial Varying Exposure (SVE) system for capturing HDR image is proposed.