A Real-Time Content Adaptation Framework for Exploiting ROI Scalability in H.264/AVC

Efficient Region-of-Interest Scalable Video Coding with Adaptive Bit-Rate Control

“…The algorithm searches, in the current frame, a region with a color histogram similar to the histogram of the ROI from the previous frame. Second, the compressed domain approach exploits DCT coefficients, motion vectors [12][13][14], or mode decisions instead of original pixel data as resources, in order to reduce computational complexity of ROI detection and tracking. The encoded data is not credible enough or insufficient to detect and track moving objects.…”

“…According to [27], the proposed ROI-based compression achieves better perceptual quality and improves coding efficiency. Moreover, [12] relates to extracting the ROIs (i.e., of an original bitstream by introducing a description-driven content adaptation framework). According to [12], two methods for the ROI extraction are implemented: (a) a removal of non-ROI portions of a bitstream and (b) a replacement of the coded background with corresponding placeholder slices.…”

Understanding User Experience of Mobile Video: Framework, Measurement, and Optimization

“…Jerbi, Wang and Shirani (2005) have adopted a non-linear transformation to duplicated ROI macroblocks before encoding, and an inverse non-linear transformation after decoding, in which way a high transmission robustness can be achieved in an error-prone channel. Another approach is to encode the ROI region separately from the background and transmit it with a higher priority to ensure the most important information will not be lost via the error-prone channel (Chen, Song, Yang & Zhang, 2007;Lambert et al, 2006).…”

Recent Advances in Region-of-interest Video Coding

“…The bit-rate control has been intensively studied in existing single layer coding standards, such as MPEG 2, MPEG 4, and H.264/AVC (Li et al, 2003). According to the existing single layer rate control schemes, the encoder employs the rate control as a way to control varying bit-rate characteristics of the coded bit-stream.…”

“…b. Allocate the remaining bits to all non-coded macroblocks (MBs) for each predefined slice in the current frame of the particular SVC layer. c. Estimate the MAD (Mean Absolute Difference) for the current macroblock in the current slice by a linear prediction model (Li et al, 2003;Lim et al, 2005) using the actual MAD of the macroblocks in the co-located position of the previous slices (in the previous frames) within the same SVC layer and the MAD of neighbor macroblocks in the current slice. d. Estimate a set of groups of coding modes (e.g., modes such as Inter-Search16X8, InterSearch8X16, Inter-Search8X8, Inter-Search8X4, Inter-Search4X8, Inter-Search4X4 modes, and the like) of the current macroblock in the current frame within the above SVC layer by using the actual group of coding modes for the macroblocks in the co-located positions of the previous frame(s) and the actual group of coding modes of neighbor macroblocks in the current frame.…”