In recent years, the advance in computer, IC, internet, cloud, and multimedia technologies have led to widespread usage of digital photos and videos in our everyday multimedia-enabled lives. A consumer can easily have many photo albums each of which contain many JPEG images taken in the same occasion. Typically many photos taken in the same occasion are similar, with similar people and background. With explosive growth of photo albums in the consumers personal computer and in the cloud (such as facebook, twitter, Google, Renren, etc), the cost to store and transmit such digital images can be very significant. Thus it is important to compress these images efficiently. In this thesis, we propose some novel effective photo album compression algorithms to achieve higher compression efficien...[ Read more ]

In recent years, the advance in computer, IC, internet, cloud, and multimedia technologies have led to widespread usage of digital photos and videos in our everyday multimedia-enabled lives. A consumer can easily have many photo albums each of which contain many JPEG images taken in the same occasion. Typically many photos taken in the same occasion are similar, with similar people and background. With explosive growth of photo albums in the consumers personal computer and in the cloud (such as facebook, twitter, Google, Renren, etc), the cost to store and transmit such digital images can be very significant. Thus it is important to compress these images efficiently. In this thesis, we propose some novel effective photo album compression algorithms to achieve higher compression efficiency than existing methods.

Observing that many images in albums are similar, we assume that similar images can be clustered and identified and we develop efficient ways to compress each set of similar JPEG images using video coding techniques. Our approach is to arrange all the similar images into some kind tree structure with many roots and branches and then apply video coding technique along each branch. To maximize the inter-frame correlation between adjacent photos along the branches in a tree, we consider both the minimum spanning tree (MST) and the spanning forest to achieve minimum inter-frame prediction cost. As an unconstrained tree can be very deep which can result in very long delay during random access, we consider trees with limited depth to achieve good random access performance. Taking advantage of the latest High Efficiency Video Coding (HEVC) standard, we consider several possible HEVC frame configuration in search of the best. Our methods can support common photo albums operations such as addition of new images, deletion of useless images and modification of existing images. Experiments suggest that the proposed methods can achieve signficant gain in coding efficiency compared with the common JPEG format.

Apart from the above research, this thesis also contains an improvement to HEVC. In the state-of-art video coding standard HEVC, temporal motion vector (MV) predictor is adopted in order to improve coding efficiency. However, motion vector information in reference frames, which is used by temporal MV predictor, takes large amount of bits in memory storage. Therefore motion data needs to be compressed before storing into buffer. Accordingly we propose an adaptive motion data storage reduction method. First, it divides the current 16x16 block in the reference frame into four partitions. One MV is sampled from each partition and all sampled MVs form a MV candidate set. Then it will check if one or two MVs should be stored into the MV buffer by checking the maximum distance between any two of the MVs in the candidate set. If the maximum distance is greater than a certain threshold, the motion data of the two MVs that have maximum distance are put into memory; otherwise the motion data of the upper left block is stored. The basic goal of the proposed method is to improve the accuracy of temporal MV predictor at the same time reducing motion data memory size. Simulation results show that compared to the original HEVC MV memory compression method in the 4th JCT-VC meeting, the proposed scheme achieves a coding gain of 0.5% to 0.6%; and the memory size is reduced by more than 87.5% comparing to without using motion data compression.