1 Department of Photonics Engineering, Technical University of Denmark2 Coding and Visual Communication, Department of Photonics Engineering, Technical University of Denmark
Post processing of decoded video and local backlight dimming for LCD technology with LED-based backlight
The quality of digital images and video signal on visual media such as TV screens and LCD displays is affected by two main factors; the display technology and compression standards. Accurate knowledge about the characteristics of display and the video signal can be utilized to develop advanced algorithms for signal (image or video) enhancement. One particular application of such algorithms is the case of LCDs with dynamic local backlight. The thesis addressed two main problems; first, designing algorithms that improve the visual quality of perceived image and video and reduce power consumption on local LED-LCD backlight. Second, removing the digital video codec artifacts such as blocking and ringing artifacts by post-processing algorithms. A novel algorithm based on image features with optimal balance between visual quality and power consumption was developed. In addition, to remove flickering artifact in LED-LCD display, an adaptive filter was proposed. Several experiments were performed to test and compare the proposed algorithms with the State of the Art algorithms. Different types of visual artifacts may occur in a video content shown on a display. They can be spatial artifacts such as blocking and ringing or temporal artifacts such as flickering or mosquito. In this thesis, some artifact reduction algorithms for MPEG-2 and H.264/AVC video are proposed.