Journal of South China University of Technology(Natural Science Edition)

Large-Capacity Lossless HEVC Information Hiding Based on Index Parameter Modification

HU Yongjian GONG Wenbin LIU Beibei LIU Shuowei ZHU Meineng

2018, 46(5): 1-8. doi:10.3969/j.issn.1000-565X.2018.05.001

Asbtract ( 298 )

PDF (2327KB) ( 424 )

References | Related Articles | Metrics

Traditional video information hiding usually modifies intraframe prediction modes，motion vectors or prediction residual coefficients for secrete message embedding． For HEVC ( High Efficiency Video Coding) standard，however， it adopts an advanced motion vector prediction ( AMVP) which uses both spatial motion vector and temporal motion vector of the adjacent block to predict the motion vector of the current block． The motion vector index is given to a prediction motion vector when it is used． Due to this new technique， this article first proposes to embed the secret information by modifying the index value． In order to guarantee the quality of the modified video， we analyze the relationship among the index value，the predicted motion vector and the motion vector residual so that we can accordingly modify the corresponding motion vector residual． This motion vector can remain unchanged after embedding． We control the capacity and bit rate change by means of a threshold． Experimental results show that our method has a larger embedded capacity and smaller bit rate increase． In particular，our method can ensure that the quality of embedded video is the same as the original video．

JPEG Image Steganography Based on Side Information Estimation

WANG Lina WANG Bo ZHAI Liming XU Yibo

2018, 46(5): 9-15. doi:10.3969/j.issn.1000-565X.2018.05.002

Asbtract ( 290 )

PDF (573KB) ( 317 )

References | Related Articles | Metrics

JPEG steganography usually takes advantage of the side information of raw images to improve steganographic security． However， the raw images are indispensable for existing JPEG steganography to apply side information． When the raw images are not available， the side information cannot be used to design more secure steganography． In this paper，we propose to improve the steganographic security by estimating the side information from JPEG images when the raw images are not available． Firstly， the convolution neural networks are used to restore the estimated spatial images from JPEG images． Secondly，the rounding errors are calculated with the estimated images． Then， the rounding errors are utilized to control the modification directions of DCT coefficients in the embedding process． Finally，by integrating the selected modification directions and the distortion costs of DCT coefficients， the STC-based steganographic embedding is implemented． The proposed method can apply to existing adaptive JPEG steganographic schemes，such as JC-UED and J-UNIWAＲD． The experimental results show that the proposed method can effectively estimate the side information from JPEG images and can improve the security of existing adaptive JPEG steganographic schemes．

A Reversible Visual Transformation Algorithm for JPEG Images

WU Haotian TANG Hongxiang WANG Jianming

2018, 46(5): 16-21. doi:10.3969/j.issn.1000-565X.2018.05.003

Asbtract ( 287 )

PDF (1976KB) ( 273 )

References | Related Articles | Metrics

Image reversible visual transformation is an emerging image processing technique that can exactly recover the original image from the visually transformed image． It can be used in the applications of secret image encryption and transmission． A new image reversible visual transformation algorithm is proposed for JPEG images with low complexity． Since JPEG format is one of the most used image types in the daily usages， study of data hiding and reversible visual transformation of JPEG images is important and attracts much attention from the community． The proposed algorithm replaces the last four DCT coefficients to embed the transform information． Consequently， the original image can be restored without any distortion by extracting the hidden data． The experimental results show that the proposed algorithm can change the secret image content while achieving reversible image visual transformation． That means the original JPEG image can be exactly recovered from the corresponding visually transformed one．

AMR steganalysis algorithm based on intra group correlation of pitch delay

REN Yanzhen LIU Dengkai YANG Jing WANG Lina

2018, 46(5): 22-31. doi:10.3969/j.issn.1000-565X.2018.05.004

Asbtract ( 261 )

PDF (1859KB) ( 355 )

References | Related Articles | Metrics

As a standard of speech compression coding for mobile Internet，AMＲ has been widely used，and also provides a new carrier for steganography． Due to the inaccurate prediction of pitch delay parameters，the existing steganographic algorithms hide the information by adjusting the pitch delay parameters． In this paper，we find the correlation between the pitch delay of each subframe by analyzing the characteristics of the pitch delay prediction coding of AMＲ encoding algorithm． Thus，we propose a new steganalysis algorithm which consists of features based on first-order Markov transition probability in subframe group and the second-order differential Markov transition probability feature of pitch delay． The experimental results show that， in the mixed training environment， the accuracy of the proposed algorithm is significantly improved，compared with the existing algorithms，especially in the case of low embedding rate， the performance is improved significantly． In the case of 10% relative embedding rate， the detection accuracy of stego samples is improved by 1% ～10% compared with the existing algorithms．

Coverless Information Hiding Algorithm Based on Image Coding and Stitching

WU Jianbin JIA Yanke LIU Yiwen

2018, 46(5): 32-38. doi:10.3969/j.issn.1000-565X.2018.05.005

Asbtract ( 228 )

PDF (1251KB) ( 305 )

References | Related Articles | Metrics

In this paper，a coverless information hiding method is introduced on the basis of image coding and splicing in order to improve effectively the security and efficiency of covert communication． Firstly， the image is divided into blocks，and the entropy of each block is calculated to construct an entropy matrix． The entropy matrix is quantized and then the mapping relation is constructed between the random number sequence and the image feature． Secondly， the random number sequence to be represented is divided into segments，and the correct images are extracted from the complete image database corresponding to the information segments． Finally，those selected images are spliced with a certain positional relationship． In this way，the stitching image is a secret carrier for transmission，which makes a high-efficiency covert communication with a coverless information hiding． Besides， the secret information is encoded with CＲC codes to improve the security． The experimental results show that the proposed algorithm has great tolerance of JPEG compression attack and salt ＆ pepper noise attack，and has a good performance of antisteganalysis． This method is significantly valuable in applying covert communication with low capacity and high security．

Large-Scale JPEG Image Steganalysis Based on DRN

TAN Shunquan LIU Guangqing ZENG Jishen LI Bin

2018, 46(5): 39-46. doi:10.3969/j.issn.1000-565X.2018.05.006

Asbtract ( 265 )

PDF (1381KB) ( 326 )

References | Related Articles | Metrics

The traditional steganalysis applies Ｒich Model features through Ensemble Classifier to achieve high detection performance． While the deep learning framework shows more powerful detection performance than traditional ones in steganalysis so far． It has been shown that the deep residual network is similar to the ensemble classifier． To confirm whether or not Xu's network，based on the steganalyzer of deep residual network which we find is not deep enough，is characteristic of the features mentioned above，we introduce deeper bottleneck architecture and reproduction of building blocks to expand them respectively，and we get four variants—bottleneck network，30layer ＲesNet，40layer ＲesNet and 50layer ＲesNet． In this article，three experiments are introduced． The first is to train the Xu's network and the variants in order to obtain the optimal models． As a result，we found that the performance of deeper network is not better than that of Xu's network． The second is to remove a building block， proving that the path in the residual network does not depend on each other． The third is to re-order some building blocks，indicating that the residual network to a certain extent can be re-configured． Finally we conclude that Xu's network is also similar to ensembles of relatively shallow networks．

Review of Physical-Layer Authentication Key Technology in Wireless Communication

Xu Zhengguang

2018, 46(5): 47-52. doi:10.3969/j.issn.1000-565X.2018.05.007

Asbtract ( 321 )

PDF (366KB) ( 740 )

References | Related Articles | Metrics

Physicallayer authentication technology is a new type of identity authentication in recent years． It embeds identity authentication information in the wireless communication signal of physical layer， so as to realize identity authentication in physical layer． On the one hand， this technology can achieve authentication，with transparency of upper layer protocol，and lower overhead of upper layer protocol; on the other hand， it can avoid identity authentication attacks for highlevel protocols and improve system security． This paper introduces two key technologies of physical layer authentication: radiofrequency fingerprint and radiofrequency watermarking． Then，a unified classification system for the proposed methods in the present literature is established， the advantages and disadvantages of the existing schemes are investigated，and，finally，the future development direction in this field is discussed．

Image Hashing Authentication Method Based on

Deep Convolution Neural Network

JIANG Cuiling PANG Yilin LIN Jiajun KANG Zhoumao

2018, 46(5): 53-59. doi:10.3969/j.issn.1000-565X.2018.05.008

Asbtract ( 357 )

PDF (2098KB) ( 388 )

References | Related Articles | Metrics

This paper presents a scheme of deep convolution network for image hashing authentication． First，the AlexNet model of deep convolution network is constructed and the given network performance is achieved through training． Then， the trained network is used to extract image features and generate image-hashing series for content authentication． The experimental results show that in comparison with other methods，the proposed method has a higher discrimination and an acceptable robustness against content-preserving operations such as random attack， rotation，JPEG compression，and additive Gaussian noise． Receiver operating characteristics ( ROC) curve comparison demonstrates that the proposed method is able to attain a desirable compromise between the robustness and discrimination．

Digital Video Frame Duplication Detection Based on TCS-LBP Operator

LI Meng HU Yongjian LIU Beibei LIU Haiyong LIAO Guangjun

2018, 46(5): 60-67,99. doi:10.3969/j.issn.1000-565X.2018.05.009

Asbtract ( 286 )

PDF (1906KB) ( 367 )

References | Related Articles | Metrics

A spatial and temporal feature-based detection algorithm is proposed for surveillance video duplication detection． Motivated by the design of LBP ( Local Binary Patterns) operator，we design a temporal domain CenterSymmetric Local Binary Pattern ( TCS-LBP) operator which can reflect the relationship between the pixel on the current frame and those pixels on its front and back frames in the same spatial location． By calculating the TCS-LBP value pixel-by-pixel on the current frame，we construct its correspondent feature image which can reflect both spatial and temporal features of the current frame． The feature image is thus used for detecting video duplication framebyframe． For the obtained results，we propose schemes for false alarms removal and missing detection correction． Finally，we accurately determine the boundaries of forgery region． Experimental results demonstrate that the proposed algorithm has good performance． It well outperforms the two similar algorithms in the literature．

Watermarking Algorithm for Tamper Detection and Recovery of EInvoice Proof Authentication

YIN Bangxu CHEN Fan HE Hongjie

2018, 46(5): 68-77. doi:10.3969/j.issn.1000-565X.2018.05.010

Asbtract ( 367 )

PDF (2006KB) ( 365 )

References | Related Articles | Metrics

By taking into consideration the detection performance， recovery quality and file size， this paper proposes an authentication-watermarking algorithm for an einvoice proof image in GIF format． According to the characteristics of an einvoice image invoice，all blocks are classified into an important block and an unimportant one． The authentication watermark of each block is generated according to the type code and content itself，and is inserted the least significant bit ( LSB) itself． To improve imperceptibility and filesize， the recovery watermark with unfixed length of each important block is generated and is embedded in the many unimportant blocks randomly chosen on the basis of secret key． The validity of an einvoice image is detected by threelayered detection by means of authentication watermark， type code and recovery watermark． The watermark capacity of the proposed scheme is reduced to about 0. 5 bpp ( bit per pixel) since the important blocks are used to generate the recovery watermark with unfixed length． The low watermark payload not only improves the quality of watermarked einvoice image，but also reduces the file size of watermarked GIF image． Furthermore， the tamper detection performance is improved due to the threelayered detection． Compared with the current similar watermarking methods，the watermarked einvoice image generated by the proposed scheme has a better quality，which means that its filesize is half in comparison with that generated by the current methods，without influence of the tamper detection and recovery performance．

Review of Image Ｒetrieval Technologies in Encrypted Domain

MEI Yuan YE Dengpan LIU Changrui

2018, 46(5): 78-86. doi:10.3969/j.issn.1000-565X.2018.05.011

Asbtract ( 264 )

PDF (859KB) ( 480 )

References | Related Articles | Metrics

With the growing popularity of the cloud storage，more and more enterprises or individuals having huge amounts of multimedia data can use cheap storage space and a wide variety of access methods． Considering the privacy security，before multimedia data are uploaded into the cloud，encryption is the simplest and the most direct method． However，most applications do not support managed and operated encrypted data． Therefore，how to effectively manage and retrieve the encrypted data becomes a challenging problem． Because of this problem， this paper first proposes a systematic framework，analyzes and summarizes several key technologies to search images in encrypted domain including encrypted technology，retrieval technology for image features，technology for safe index， etc. ． Then，several representative retrieval algorithms and their development are analyzed and compared． Finally， the future of the technology to search encrypted images is prospected．With the growing popularity of the cloud storage，more and more enterprises or individuals having huge amounts of multimedia data can use cheap storage space and a wide variety of access methods． Considering the privacy security，before multimedia data are uploaded into the cloud，encryption is the simplest and the most direct method． However，most applications do not support managed and operated encrypted data． Therefore，how to effectively manage and retrieve the encrypted data becomes a challenging problem． Because of this problem， this paper first proposes a systematic framework，analyzes and summarizes several key technologies to search images in encrypted domain including encrypted technology，retrieval technology for image features，technology for safe index， etc. ． Then，several representative retrieval algorithms and their development are analyzed and compared． Finally， the future of the technology to search encrypted images is prospected．

Experimental Investigation into Emission Characteristics of HCl During Co-Combustion of Wood Sawdust and PVC

#br#

MA Xiaoqian ZENG Xianghao TANG Yuting DAI Minquan

2018, 46(5): 87-92. doi:10.3969/j.issn.1000-565X.2018.05.012

Asbtract ( 251 )

PDF (1653KB) ( 375 )

References | Related Articles | Metrics

types of absorber，atmosphere， ratio of calcium to chlorine，and fuel mixture ratio on the dechlorination of wood chips and PVC during cofiring under O2 /CO2 atmosphere were investigated by using a tube furnace． The results show that with the increment of the temperature， the dechlorination efficiency increases firstly and decreases subsequently． Meanwhile， the highest dechlorination efficiency is 71. 4% at 700℃． The ratio of additives has a certain impact on dechlorination． Modified calcium carbonate can increase the dechlorination rate at a high temperature，because the pore structure becomes more abundant and the specific surface area becomes relatively larger． As the mixing ratio of sawdust rises，the dechlorination efficiency increases， and the dechlorination efficiency in O2 /CO2 atmosphere is higher than that in O2 /N2 atmosphere． The analysis of BET and SEM/EDS on sorbent and product indicates that the surface of the product particle covers the layer of dense CaCl2，which prevents the HCl from penetrating the particle surface into the dechlorination agent inside the full reaction． The reaction becomes inadequate，which leads to the reduction in dechlorination efficiency．

Control Strategy Considering Power Coordinated Control for Low-Voltage Ride Through Two-Stage PV Inverter

OUYANG Sen MA Wenjie

2018, 46(5): 93-99. doi:10.3969/j.issn.1000-565X.2018.05.013

Asbtract ( 338 )

PDF (1626KB) ( 254 )

References | Related Articles | Metrics

In order to enhance the low-voltage ride capability of the twostage grid-connected system PV inverter based on Boost circuit，a control strategy with the DC bus voltage control loop based on Boost for LVＲT is designed in this paper． Firstly， the conventional Boost voltage control mode and output current control strategy of the twostage PV inverter are analyzed． A DC bus voltage control outer loop based on PV array maximum power point voltage feed forward is introduced to change the Boost voltage control mode． In this way，the purposes of making the DC bus voltage stable and preventing over current of the inverter are to be achieved． At the same time，the reference value of the active current is adjusted according to the voltage drop depth． On the premise that the inverter is not over current， the maximum reactive current is provided to help the power grid recovery． Finally，a 100kW PV inverter PSCAD simulation model is built to verify the proposed control strategy． The result shows that the method not only can suppress the DC bus voltage rise rapidly and effectively，but also can make the inverter provide more dynamic reactive power support for the grid even when the voltage drops．

Joint Optimization of Adaptive Ｒobust Active and Ｒeactive Power in Active Distribution Network#br#

DONG Ping SUN Xinglu WANG Yaping LIN Yun

2018, 46(5): 100-108,116. doi:10.3969/j.issn.1000-565X.2018.05.014

Asbtract ( 254 )

PDF (1525KB) ( 432 )

References | Related Articles | Metrics

Considering the influence of the randomness of the distributed power supply on the operation of the active distribution network ( ADN) ，an adaptive robust optimization ( AＲO) method is proposed and combined with active and reactive power optimization for ADN to improve system performance． First of all，considering all kinds of discrete and continuous adjustable devices，an ADN operation optimization model based on Distflow power flow model is established; then， the AＲO model is used to deal with the model in consideration of the randomness of the PV output and the column-and-constraint generation algorithm is used to deal with the model; an improved outer approximation algorithm is proposed to solve the dual problem of nonlinearity． In addition， the battery energy storage system ( BESS) is used as the realtime controllable adjustment，and applied to the second stage of AＲO method． Finally， the extended IEEE33-node and PG＆E69-node system are used to perform the simulation calculation，the deterministic method and the method of the first stage control device used only as BESS are compared and analyzed， and the validity and reliability of the method of this paper which is used to improve the operation ability of ADN are thus verified．

Control Strategy of T-Type Three-Level Inverter Based on Integrated Dead-Time Elimination and Compensation

HU Bihua KANG Longyu CHENG Jiancai WANG Zefeng WU Jingyue

2018, 46(5): 109-116. doi:10.3969/j.issn.1000-565X.2018.05.015

Asbtract ( 375 )

PDF (2316KB) ( 485 )

References | Related Articles | Metrics

In Ttype converters， to avoid the power switches of the same leg to conduct simultaneously during altering the power switches， the deadtime has to be inserted into the switching sequence． However， the inserted deadtime increases the current THD and reduces the output voltage and current． In this paper，an approach which integrates deadtime elimination and compensation is proposed to remove the adverse effect of deadtime based on the function of the deadtime． When the phase current is much larger than the phase current ripple，the deadtime elimination is applied to avoid short circuit without dead time． When the phase current is near zero-crossing point， the deadtime elimination is inapplicable because of the disturbance of current ripples，and the deadtime must be used to improve the stability of inverter． Therefore， the deadtime compensation is used to remove the adverse effect of deadtime． At last， the experimental results demonstrate that the proposed approach is an effective solution to the problems caused by deadtime

A Method to Analyze and Optimize Assembly Performance of Outer Seals of Car Window Belt-Line

LI Min TAN Meng SHANGGUAN Wenbin LI Zhichao LIU Jie YOU Limin

2018, 46(5): 117-124. doi:10.3969/j.issn.1000-565X.2018.05.016

Asbtract ( 211 )

PDF (1339KB) ( 467 )

References | Related Articles | Metrics

Outer seals of car window beltline play an important role in waterproof，noise insulation and decoration． The assembly and compression performance is an important factor for the evaluation of the design of the outer seals of car window beltline． In this paper，a calculation method for estimating the assembly performance of the outer seals of car window beltline is proposed，and assembly performance parameters are calculated and compared with measured values． The results show that the simulation and calculation method is correct． Then， taking the length， the thickness and the interference of the clamping tooth as design variables，a parametric model of the seals is established． On this basis， the optimization method of the assembly performance of the seals is presented by using the response surface model and multi-objective particle swarm optimization．

Evaluation of Flow Field Simulation of Turbine in Water-Medium Hydrodynamic Coupling Based on PIV Experiment

2018, 46(5): 125-134. doi:10.3969/j.issn.1000-565X.2018.05.017

Asbtract ( 207 )

PDF (4892KB) ( 255 )

References | Related Articles | Metrics

In order to reveal the characteristics and evolution law of multiscale vortex flow field in water medium hydrodynamic coupling based on the CFD technique，four kinds of turbulence models ( DES，DDES，IDDES， LES) were used to simulate the internal multiscale vortex flow field of turbine in hydrodynamic coupling in braking and traction conditions． The flow images of internal flow field in turbine were collected by particle image velocimetry ( PIV) technology， static and dynamic image calibration methods were used to calculate the flow field under different conditions． The applicability of the four turbulence models was evaluated on the basis of the comparison of PIV experimental results． The results show that under the braking condition，the simulation results of multiscale vortex flow field in mainstream region used by LES turbulence model tend to be true， the flow velocity range is from 3. 52 to 3. 81m/s，and the vorticity range is from 480 to 540s －1． While the simulation results of multiscale vortex flow field in the near-wall region of the blade used by IDDES turbulence model are outstanding in performance， the flow velocity range is from 3. 14 to 3. 51m/s， in this area， simulation results of vortex field used by DES turbulence model are better than those of LES turbulence model，and the vorticity range is from 500 to 570s －1． As far as traction condition is concerned，the simulation results are distorted when DDES turbulence model and IDDES turbulence model are used，and the simulation results of multiscale vortices in mainstream region used by DES turbulence model are not as good as those of LES turbulence model，but it can reflect the basic movement trend of multiscale vortices along the circumferential direction． The simulation results used by LES turbulence model are in good agreement with the PIV experimental results， it is possible to embody the mutual mixing process of these multiscale vortices moving along the circumferential direction， the flow velocity range is from 1. 40 to 3. 60m/s，and the vorticity range is from 30 to 130s －1． It will provide some technical guidance for the precise simulation of hydrodynamic

coupling．

Development of Lightweight All-Sealed Wire Feeder for Underwater Robotic Welding System

HAN Leigang FAN Wenyan WANG Guohe WANG Zhenmin

2018, 46(5): 135-140. doi:10.3969/j.issn.1000-565X.2018.05.018

Asbtract ( 240 )

PDF (2040KB) ( 276 )

References | Related Articles | Metrics

Aiming at the repair needs of the underwater Ｒobotic MIG welding for nuclear fuel pool，a lightweighted and entirely-sealed wire feeder was developed． The overall lightweighting is achieved by means of aluminum alloy shell，glass steel panel and other devices; the double driving roller wire feeding mechanism is adopted，and the miniaturization of the device is realized through the optimized layout of the components like welding wire tray． The shell is connected with the compressed gas through a trachea to realize the self-adjustment of the pressure of the device with the change of water depth． To compensate the change of wire feeding speed，a digital driving control system based on AＲM is designed which adopted negative feedback control of armature voltage and high frequency chopping technology． The digital communication interface circuit based on CAN is used to realize the digital integration with the underwater robot． Test results of underwater welding at the level of 20 m depth show that the wire feeding device has the advantages of high shell strength，stable and uniform wire feeding，quick response and good adaptability，and thus meet the requirements of underwater welding repair．

Capacity Fading Detection of Power Battery on Electric Vehicle by Using Tomographic Image

#br#

2018, 46(5): 141-146. doi:10.3969/j.issn.1000-565X.2018.05.019

Asbtract ( 358 )

PDF (1017KB) ( 260 )

Table of Content