• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.37 no.5
• /
• pp.54-63
• /
• 2000

We present a multi-resolution block matching algorithm (BMA) for fast motion estimation At the coarsest level, a motion vector (MV) having minimum matching error is chosen via a full search, and a MV with minimum matching error is concurrently found among the MVs of the spatially adjacent blocks Here, to examine the spatial MVs accurately, we propose an efficient method for searching full resolution MV s without MV quantization even at the coarsest level The chosen two MV s are used as the initial search centers at the middle level At the middle level, the local search is performed within much smaller search area around each search center If the method used at the coarsest level is adopted here, the local searches can be done at integer-pel accuracy A MV having minimum matching error is selected within the local search areas, and then the final level search is performed around this initial search center Since the local searches are performed at integer-pel accuracy at the middle level, the local search at the finest level does not take an effect on the overall performance So we can skip the final level search without performance degradation, thereby the search speed increases Simulation results show that in comparison with full search BMA, the proposed BMA without the final level search achieves a speed-up factor over 200 with minor PSNR degradation of 02dB at most, under a normal MPEG2 coding environment Furthermore, our scheme IS also suitable for hardware implementation due to regular data-flow.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.6
• /
• pp.641-651
• /
• 2002

Recently, many advanced video application services over the mobile wireless networks have required a transcoder which can efficiently reduce the size of compressed video bitstreams. The transcoder can be worked in either the spatial domain or the DCT domain. In this paper, we propose a new fast hybrid-type transcoder which can efficiently reduce the frame size with keeping the visual quality. The proposed scheme consists of two major processes: a transform domain process and a spatial domain process. We also propose a scheme for coding mode selection and motion vector refinement. Experimental results show that our approach can reduce the computational complexity more than any other conventional spatial-domain transcoder with keeping the visual quality.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.1A
• /
• pp.51-56
• /
• 2012

In order to reduce the complexity of HEVC, we propose a fast mode decision algorithm in depth videos. Since almost CU mode is decided as SKIP mode in depth-continuity regions, we design the algorithm using the property of depth videos. If cost of SKIP is smaller than the multiplication between the threshold for EarlySKIP and average cost of SKIP, EarlySKIP is performed. Otherwise, we calculate Inter $2N{\times}2N$. Then, if motion vector of Inter $2N{\times}2N$ is 0 and variance of CU is smaller than threshold for inter, we skip Inter $2N{\times}N$, Inter $N{\times}2N$. Experimental results show that our proposed algorithm reduces the encoding time from 39% to 82% with negligible PSNR loss and bitrate increase.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.2
• /
• pp.129-137
• /
• 2001

Matching pursuit algorithm was successfully demonstrated useful in low bit-rate video coding, However, one of the practical concerns related to applying the matching pursuit algorithm to application is its massive computation required for finding bases whose weighted sum best approximates the given input image. The main contribution of this paper is that we provide a new method that can drastically reduce the computational load without any degradation of image quality. Its main idea is based on reducing the number of inner product calculation required for finding best bases because the complexity of matching pursuit algorithm is due to the exhaustive local inner product calculation. As the first step, we compute a matrix which is the 1-D inner product of the given motion-compensated error input image with the 1-D vertical Gabor functions using the separable property of Gabor bases. In the second step, we calculate length of each vector in the matrix that corresponds to 1-D horizontal Gabor function, and compare the length with the current maximum absolute inner product value so far. According to the result of this comparison, one can decide whether or not to calculate the inner product. Since most of them do not need to calculate the inner product value, one can significantly reduce the computational load. Experimental results show that proposed method reduces about 70% of inner product calculation compared to the Neff's fast algorithm without any degradation of image quality.