-
Shunsuke YAMAKI, Masahide ABE, Masayuki KAWAMATA
原稿種別: PAPER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1097-1108
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
This paper proposes the statistical analysis of phase-only correlation functions between two real signals with phase-spectrum differences. For real signals, their phase-spectrum differences have odd-symmetry with respect to frequency indices. We assume phase-spectrum differences between two signals to be random variables. We next derive the expectation and variance of the POC functions considering the odd-symmetry of the phase-spectrum differences. As a result, the expectation and variance of the POC functions can be expressed by characteristic functions or trigonometric moments of the phase-spectrum differences. Furthermore, it is shown that the peak value of the POC function monotonically decreases and the sidelobe values monotonically increase as the variance of the phase-spectrum differences increases.
抄録全体を表示
-
Taizo SUZUKI, Masaaki IKEHARA
原稿種別: PAPER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1109-1118
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
We present a lifting-based lapped transform (L-LT) and a reversible symmetric extension (RSE) in the boundary processing for more effective lossy-to-lossless image coding of data with various qualities from only one piece of lossless compressed data. The proposed dual-DCT-lifting-based LT (D2L-LT) parallel processes two identical LTs and consists of 1-D and 2-D DCT-liftings which allow the direct use of a DCT matrix in each lifting coefficient. Since the DCT-lifting can utilize any existing DCT software or hardware, it has great potential for elegant implementations that are dependent on the architecture and DCT algorithm used. In addition, we present an improved RSE (IRSE) that works by recalculating the boundary processing and solves the boundary problem that the DCT-lifting-based L-LT (DL-LT) has. We show that D2L-LT with IRSE mostly outperforms conventional L-LTs in lossy-to-lossless image coding.
抄録全体を表示
-
Takuro YAMAGUCHI, Masaaki IKEHARA
原稿種別: PAPER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1119-1126
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
Image interpolation is one of the image upsampling technologies from a single input image. This technology obtains high resolution images by fitting functions or models. Although image interpolation methods are faster than other upsampling technologies, they tend to cause jaggies and blurs in edge and texture regions. Multi-surface Fitting is one of the image upsampling techniques from multiple input images. This algorithm utilizes multiple local functions and the weighted means of the estimations in each local function. Multi-surface Fitting obtains high quality upsampled images. However, its quality depends on the number of input images. Therefore, this method is used in only limited situations. In this paper, we propose an image interpolation method with both high quality and a low computational cost which can be used in many situations. We adapt the idea of Multi-surface Fitting for the image upsampling problems from a single input image. We also utilize local functions to reduce blurs. To improve the reliability of each local function, we introduce new weights in the estimation of the local functions. Besides, we improve the weights for weighted means to estimate a target pixel. Moreover, we utilize convolutions with small filters instead of the calculation of each local function in order to reduce the computational cost. Experimental results show our method obtains high quality output images without jaggies and blurs in short computational time.
抄録全体を表示
-
Tomoko KAWASE, Kenta NIWA, Masakiyo FUJIMOTO, Kazunori KOBAYASHI, Shok ...
原稿種別: PAPER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1127-1136
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
We propose a microphone array speech enhancement method that integrates spatial-cue-based source power spectral density (PSD) estimation and statistical speech model-based PSD estimation. The goal of this research was to clearly pick up target speech even in noisy environments such as crowded places, factories, and cars running at high speed. Beamforming with post-Wiener filtering is commonly used in many conventional studies on microphone-array noise reduction. For calculating a Wiener filter, speech/noise PSDs are essential, and they are estimated using spatial cues obtained from microphone observations. Assuming that the sound sources are sparse in the temporal-spatial domain, speech/noise PSDs may be estimated accurately. However, PSD estimation errors increase under circumstances beyond this assumption. In this study, we integrated speech models and PSD-estimation-in-beamspace method to correct speech/noise PSD estimation errors. The roughly estimated noise PSD was obtained frame-by-frame by analyzing spatial cues from array observations. By combining noise PSD with the statistical model of clean-speech, the relationships between the PSD of the observed signal and that of the target speech, hereafter called the observation model, could be described without pre-training. By exploiting Bayes' theorem, a Wiener filter is statistically generated from observation models. Experiments conducted to evaluate the proposed method showed that the signal-to-noise ratio and naturalness of the output speech signal were significantly better than that with conventional methods.
抄録全体を表示
-
Toru TANZAWA
原稿種別: PAPER
専門分野: Circuit Theory
2017 年 E100.A 巻 5 号 p.
1137-1144
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
An output voltage-current equation of charge pump DC-DC voltage multiplier using diodes is provided to cover wide clock frequency and output current ranges for designing energy harvester operating at a near-threshold voltage or in sub-threshold region. Equivalent circuits in slow and fast switching limits are extracted. The effective threshold voltage of the diode in slow switching limit is also derived as a function of electrical characteristics of the diodes, such as the saturation current and voltage slope parameter, and design parameters such as the number of stages, capacitance per stage, parasitic capacitance at the top plate of the main boosting capacitor, and the clock frequency. The model is verified compared with SPICE simulation.
抄録全体を表示
-
Jing WANG, Li DING, Qiang LI, Hirofumi SHINOHARA, Yasuaki INOUE
原稿種別: PAPER
専門分野: VLSI Design Technology and CAD
2017 年 E100.A 巻 5 号 p.
1145-1155
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, a nanopower supply-insensitive complementary metal-oxide-semiconductor (CMOS) unit threshold voltage (Vth) extractor circuit is proposed. It meets the contemporary industry demand for portable devices that operate with very low power consumption and small output sensitivity. An α times Vth (αVth) extractor is also described, in which α varies continuously. Both incremental and decremental αVth voltages are obtained. A post-layout simulation results using HSPICE with CMOS 0.18um process show that the proposed unit Vth extractor consumes 265nW of power given a 1.6V power supply. Sensitivity to temperature is 0.022%/°C ranging from 0°C to 100°C. Sensitivity to supply voltage is 0.027%/V.
抄録全体を表示
-
Ngoc-Giao PHAM, Suk-Hwan LEE, Ki-Ryong KWON
原稿種別: PAPER
専門分野: Cryptography and Information Security
2017 年 E100.A 巻 5 号 p.
1156-1164
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
Nowadays, vector map content is widely used in the areas of life, science and the military. Due to the fact that vector maps bring great value and that their production process is expensive, a large volume of vector map data is attacked, stolen and illegally distributed by pirates. Thus, vector map data must be encrypted before being stored and transmitted in order to ensure the access and to prevent illegal copying. This paper presents a novel perceptual encryption algorithm for ensuring the secured storage and transmission of vector map data. Polyline data of vector maps are extracted to interpolate a spline curve, which is represented by an interpolating vector, the curvature degree coefficients, and control points. The proposed algorithm is based on encrypting the control points of the spline curve in the frequency domain of discrete cosine transform. Control points are transformed and selectively encrypted in the frequency domain of discrete cosine transform. They are then used in an inverse interpolation to generate the encrypted vector map. Experimental results show that the entire vector map is altered after the encryption process, and the proposed algorithm is very effective for a large dataset of vector maps.
抄録全体を表示
-
Lan V. TRUONG, Hirosuke YAMAMOTO
原稿種別: PAPER
専門分野: Information Theory
2017 年 E100.A 巻 5 号 p.
1165-1178
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, the posterior matching scheme proposed by Shayevits and Feder is extended to the Gaussian broadcast channel with feedback, and the error probabilities and achievable rate region are derived for this coding strategy by using the iterated random function theory. A variant of the Ozarow-Leung code for the general two-user broadcast channel with feedback can be realized as a special case of our coding scheme. Furthermore, for the symmetric Gaussian broadcast channel with feedback, our coding scheme achieves the linear-feedback sum-capacity like the LQG code and outperforms the Kramer code.
抄録全体を表示
-
Zheng DAI, Weimin SU, Hong GU
原稿種別: PAPER
専門分野: Communication Theory and Signals
2017 年 E100.A 巻 5 号 p.
1179-1187
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, we consider a coherently distributed (CD) source model. Since the CD source is characterized by four parameters: central azimuth direction-of-arrival (DOA), azimuth angular spread, central elevation DOA and elevation angular spread, the parameter estimation is normally complex. We propose an algorithm that combines the rotational invariance techniques (ESPRIT) and the generalized ESPRIT algorithm for the 2-dimensional (2D) central DOA estimation of CD sources. Using a pair of uniform circular arrays (UCAs), the proposed solution is able to obtain the central DOAs with both high accuracy and low computational complexity. The central elevation DOAs are estimated by using the rotational invariance relation between the two uniform circular sub-arrays. Based on the centrosymmetric structure of UCA, the generalized ESPRIT algorithm is then applied to estimate the central azimuth DOAs through one-dimensional searching. It is noteworthy that the central DOAs are estimated without any information of the deterministic angular distribution function (DADF). The performance of the proposed algorithm is demonstrated via computer simulations.
抄録全体を表示
-
Heungseop AHN, Seungwon CHOI
原稿種別: PAPER
専門分野: Communication Theory and Signals
2017 年 E100.A 巻 5 号 p.
1188-1196
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
The sub-blocking algorithm has been known as a core component in implementing a turbo decoder using a Graphic Processing Unit (GPU) to use as many cores in the GPU as possible for parallel processing. However, even though the sub-blocking algorithm allows a large number of threads in a given GPU to be adopted for processing a large number of sub-blocks in parallel, each thread must access the global memory with strided addresses, which results in uncoalesced memory access. Because uncoalesced memory access causes a lot of unnecessary memory transactions, the memory bandwidth efficiency drops significantly, possibly as low as 1/8 in the case of an Long Term Evolution (LTE) turbo decoder, depending upon the compute capability of a GPU. In this paper, we present a novel method for converting uncoalesced memory access into coalesced access in a way that completely recovers the memory bandwidth efficiency to 100% without additional overhead. Our experimental tests, performed with NVIDIA's Geforce GTX 780 Ti GPU, show that the proposed method can enhance the throughput by nearly 30% compared with a conventional turbo decoder that suffers from uncoalesced memory access. Throughput provided by the proposed method has been observed to be 51.4Mbps when the number of iterations and that of sub-blocks are set to 6 and 32, respectively, in our experimental tests, which far exceeds the performance of previous works implemented the Max-Log-MAP algorithm.
抄録全体を表示
-
Tetsuya MANABE, Takaaki HASEGAWA
原稿種別: PAPER
専門分野: Intelligent Transport System
2017 年 E100.A 巻 5 号 p.
1197-1205
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, the differences in navigation information design, which is important for kiosk-type pedestrian navigation systems, were experimentally examined depending on presence or absence of carriable navigation information in order to acquire the knowledge to contribute design guidelines of kiosk-type pedestrian navigation systems. In particular, we used route complexity information calculated using a regression equation that contained multiple factors. In the absence of carriable navigation information, both the destination arrival rate and route deviation rate improved. Easy routes were designed as M (17 to 39 characters in Japanese), while complicated routes were denoted as L (40 or more characters in Japanese). On the contrary, in the presence of carriable navigation information, the user's memory load was found to be reduced by carrying the same navigation information as kiosk-type terminals. Thus, the reconsideration of kiosk-type pedestrian navigation systems design, e.g., the means of presenting navigation information, is required. For example, if the system attaches importance to a high destination arrival rate, L_Carrying without regard to route complexity is better. If the system attaching importance to the low route deviation rate, M_Carrying in the case of easy routes and L_Carrying in the case of complicated routes have been better. Consequently, this paper presents the differences in the designs of pedestrian navigation systems depending on whether carriable navigation information is absent or present.
抄録全体を表示
-
Akira John SUZUKI, Kiyoshi MIZUI
原稿種別: PAPER
専門分野: Intelligent Transport System
2017 年 E100.A 巻 5 号 p.
1206-1214
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In autonomous vehicles, driving in traffic poses significant challenges in vehicle-to-vehicle (V2V) communication and ranging. Currently interest centers on enhanced V2V communication with multi-sensor and cooperative approaches. In this paper we propose a novel bidirectional Laser Radar Visible Light Bidirectional Communication Boomerang System (LRVLB-ComBo). LRVLB-ComBo affords nuanced real-time two-way V2V communication as a basis for complex but reliable decision-making. Our approach involves combining existing automotive laser radar with visible light boomerang systems using THSS techniques. System simulations were performed using a random mix of extraneous interference pulse to evaluate system sensitivity to noise. Results suggest that LRVLB-ComBo is a viable two-way V2V communication system with increased ranging accuracy, enabling provision of detailed bidirectional data exchange for ITS precision, energy efficiency and safety.
抄録全体を表示
-
Xina CHENG, Norikazu IKOMA, Masaaki HONDA, Takeshi IKENAGA
原稿種別: PAPER
専門分野: Vision
2017 年 E100.A 巻 5 号 p.
1215-1225
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
Significant challenges in ball tracking of sports analysis by computer vision technology are: 1) accuracy of estimated 3D ball trajectory under difficult conditions; 2) external forces added by players lead to irregular motions of the ball; 3) unpredictable situations in the real game, i.e. the ball occluded by players and other objects, complex background and changing lighting condition. With the goal of multi-view 3D ball tracking, this paper proposes an abrupt motion adaptive system model, an anti-occlusion observation model, and a spatial density-based automatic recovery based on particle filter. The system model combines two different system noises that cover the motion of the ball both in general situation and situation subject to abrupt motion caused by external force. Combination ratio of these two noises and number of particles are adaptive to the estimated motion by weight distribution of particles. The anti-occlusion observation model evaluates image feature of each camera and eliminates influence of the camera with less confidence. The spatial density, which is calculated based on 3D ball candidates filtered out by spatial homographic relationship between cameras, is proposed for generating new set of particles to recover the tracking when tracking failure is detected. Experimental results based on HDTV video sequences (2014 Inter High School Men's Volleyball Games, Japan), which were captured by four cameras located at each corner of the court, show that the success rate achieved by the proposals of 3D ball tracking is 99.42%.
抄録全体を表示
-
Yuan CHEN, Long-Ting HUANG, Xiao Long YANG, Hing Cheung SO
原稿種別: LETTER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1226-1230
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
Variance analysis is an important research topic to assess the quality of estimators. In this paper, we analyze the performance of the least ℓp-norm estimator in the presence of mixture of generalized Gaussian (MGG) noise. In the case of known density parameters, the variance expression of the ℓp-norm minimizer is first derived, for the general complex-valued signal model. Since the formula is a function of p, the optimal value of p corresponding to the minimum variance is then investigated. Simulation results show the correctness of our study and the near-optimality of the ℓp-norm minimizer compared with Cramér-Rao lower bound.
抄録全体を表示
-
Kang WU, Yijun CHEN, Huiling HOU, Wenhao CHEN, Xuwen LIANG
原稿種別: LETTER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1231-1235
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this letter, a new and accurate frequency estimation method of complex exponential signals is proposed. The proposed method divides the signal samples into several identical segments and sums up the samples belonging to the same segment respectively. Then it utilizes fast Fourier transform (FFT) algorithm with zero-padding to obtain a coarse estimation, and exploits three Fourier coefficients to interpolate a fine estimation based on least square error (LSE) criterion. Numerical results show that the proposed method can closely approach the Cramer-Rao bound (CRB) at low signal-to-noise ratios (SNRs) with different estimation ranges. Furthermore, the computational complexity of the proposed method is proportional to the estimation range, showing its practical-oriented ability. The proposed method can be useful in several applications involving carrier frequency offset (CFO) estimation for burst-mode satellite communications.
抄録全体を表示
-
Huan HAO, Huali WANG, Wanghan LV, Liang CHEN
原稿種別: LETTER
専門分野: Digital Signal Processing
2017 年 E100.A 巻 5 号 p.
1236-1239
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
This paper proposes an effective continuous super-resolution (CSR) algorithm for the multipath channel estimation. By designing a preamble including up-chirp and down-chirp symbols, the Doppler shift and multipath delay are estimated jointly by using convex programming. Simulation results show that the proposed CSR can achieve better detection probability of the number of multipaths than the eigenvalue based methods. Moreover, compared with conventional super-resolution techniques, such as MUSIC and ESPRIT methods, the proposed CSR algorithm demonstrates its advantage in root mean square error of the Doppler shift and multipath delay, especially for the closely located paths within low SNR.
抄録全体を表示
-
Zhanhu HU, Wang HU, Zhiping WANG
原稿種別: LETTER
専門分野: Systems and Control
2017 年 E100.A 巻 5 号 p.
1240-1243
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
To improve the quality of waveforms and achieve a high input power factor (IPF) for matrix rectifier, a novel quasi sliding mode control (SMC) with adaptive compensation is proposed in this letter. Applying quasi-SMC can effective obviate the disturbances of time delay and spatial lag, and SMC based on continuous function is better than discontinuous function to eliminate the chattering. Furthermore, compared with conventional compensation, an adaptive quasi-SMC compensation without any accurate detection for internal parameters is easier to be implementated, which has shown a superior advance. Theoretical analysis and experiments are carried out to validate the correctness of the novel control scheme.
抄録全体を表示
-
Tomoyuki SASAKI, Hidehiro NAKANO, Arata MIYAUCHI, Akira TAGUCHI
原稿種別: LETTER
専門分野: Nonlinear Problems
2017 年 E100.A 巻 5 号 p.
1244-1247
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, we propose a new paradigm of deterministic PSO, named piecewise-linear particle swarm optimizer (PPSO). In PPSO, each particle has two search dynamics, a convergence mode and a divergence mode. The trajectory of each particle is switched between the two dynamics and is controlled by parameters. We analyze convergence condition of each particle and investigate parameter conditions to allow particles to converge to an equilibrium point through numerical experiments. We further compare solving performances of PPSO. As a result, we report here that the solving performances of PPSO are substantially the same as or superior to those of PSO.
抄録全体を表示
-
Seokhyun SON, Myoungjin KIM
原稿種別: LETTER
専門分野: Graphs and Networks
2017 年 E100.A 巻 5 号 p.
1248-1253
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this letter, we propose a Hadoop-based Video Transcoding System (HVTS), which is designed to run on all major cloud computing services. HVTS is highly adapted to the structure and policies of Hadoop, thus it has additional capacities for transcoding, task distribution, load balancing, and content replication and distribution. To evaluate, our proposed system, we carry out two performance tests on our local testbed, transcoding and robustness to data node and task failures. The results confirmed that our system delivers satisfactory performance in facilitating seamless streaming services in cloud computing environments.
抄録全体を表示
-
Ming LI, Yupeng JIANG, Dongdai LIN, Qiuyan WANG
原稿種別: LETTER
専門分野: Cryptography and Information Security
2017 年 E100.A 巻 5 号 p.
1254-1256
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
We regard a De Bruijn sequence of order n as a bijection on $\mathbb{F}_2^n$ and consider the transition mappings between them. It is shown that there are only two conjugate transformations that always transfer De Bruijn sequences to De Bruijn sequences.
抄録全体を表示
-
Qing WU, Leyou ZHANG, Jingxia ZHANG
原稿種別: LETTER
専門分野: Cryptography and Information Security
2017 年 E100.A 巻 5 号 p.
1257-1261
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
Fuzzy techniques can implement the fine-grained access control of encrypted data in the Cloud because they support error-tolerance. In this system, using biometric attributes such as fingerprints, faces and irises as pubic parameters is advantageous over those systems based on Public Key Infrastructure (PKI). This is because biometric information is unique, unforgettable and non-transferable. However the biometric-attribute measurements are noisy and most of the existing encryption systems can not support the biometric-attribute encryption. Additionally, the previous fuzzy encryption schemes only achieve the selective security which is a weak security model. To overcome these drawbacks, we propose a new fuzzy encryption scheme based on the lattice in this letter. The proposed scheme is based on a hierarchical identity-based encryption with fixed-dimensional private keys space and thus has short public parameters and short private keys, which results in high computation efficiency. Furthermore, it achieves the strong security, i.e., adaptive security. Lastly, the security is reduced to the learning with errors (LWE) problem in the standard model.
抄録全体を表示
-
Vamoua YACHONGKA, Hideki YAGI
原稿種別: LETTER
専門分野: Information Theory
2017 年 E100.A 巻 5 号 p.
1262-1266
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
The biometrical identification system, introduced by Willems et al., is a system to identify individuals based on their measurable physical characteristics. Willems et al. characterized the identification capacity of a discrete memoryless biometrical identification system from information theoretic perspectives. Recently, Mori et al. have extended this scenario to list-decoding whose list size is an exponential function of the data length. However, as the data length increases, how the maximum identification error probability (IEP) behaves for a given rate has not yet been characterized for list-decoding. In this letter, we investigate the reliability function of the system under fixed-size list-decoding, which is the optimal exponential behavior of the maximum IEP. We then use Arimoto's argument to analyze a lower bound on the maximum IEP with list-decoding when the rate exceeds the capacity, which leads to the strong converse theorem. All results are derived under the condition that an unknown individual need not be uniformly distributed and the identification process is done without the knowledge of the prior distribution.
抄録全体を表示
-
Yu Min HWANG, Yuchan SONG, Kwang Yul KIM, Yong Sin KIM, Jae Seang LEE, ...
原稿種別: LETTER
専門分野: Communication Theory and Signals
2017 年 E100.A 巻 5 号 p.
1267-1270
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this paper, we propose a non-cooperative line-of-sight (LOS)/non-LOS channel identification algorithm with single node channel measurements based on time-of-arrival statistics. In order to improve the accuracy of channel identification, we adopt a recalibration interval in terms of measured distance to the proposed algorithm. Experimental results are presented in terms of identification probability and recalibration interval. The proposed algorithm involves a trade-off between the channel identification quality and the recalibration rate. However, depending on the recalibration interval, it is possible to greatly improve the sensitivity of the channel identification system.
抄録全体を表示
-
Yong-An JUNG, Yung-Lyul LEE, Hyoung-Kyu SONG, Young-Hwan YOU
原稿種別: LETTER
専門分野: Mobile Information Network and Personal Communications
2017 年 E100.A 巻 5 号 p.
1271-1273
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
In this letter, a simple and robust synchronization algorithm for second generation terrestrial digital video broadcasting (DVB-T2) receivers is proposed. In the proposed detection scheme, the coarse symbol timing is estimated by decimating a correlation output to give a sharper peak timing detection metric. Such a design can improve the timing synchronization accuracy as well as enhance its robustness to frequency selective channels.
抄録全体を表示
-
Jia QIN, Huihui BAI, Mengmeng ZHANG, Yao ZHAO
原稿種別: LETTER
専門分野: Image
2017 年 E100.A 巻 5 号 p.
1274-1278
発行日: 2017/05/01
公開日: 2017/05/01
ジャーナル
認証あり
High Efficiency Video Coding (HEVC) is the latest coding standard. Compared with Advanced Video coding (H.264/AVC), HEVC offers about a 50% bitrate reduction at the same reconstructed video quality. However, this new coding standard leads to enormous computational complexity, which makes it difficult to encode video in real time. Therefore, in this paper, aiming at the high complexity of intra coding in HEVC, a new fast coding unit (CU) splitting algorithm is proposed based on the decision tree. Decision tree, as a method of machine learning, can be designed to determine the size of CUs adaptively. Here, two significant features, Just Noticeable Difference (JND) values and coding bits of each CU can be extracted to train the decision tree, according to their relationships with the CUs' partitions. The experimental results have revealed that the proposed algorithm can save about 34% of time, on average, with only a small increase of BD-rate under the “All_Intra” setting, compared with the HEVC reference software.
抄録全体を表示