4 65 M = C p + f 1 C p f Sobel 1 C p f n C n 1 1 C n = B q n whn B n n w h q n F Q 13 7 F Q = a 1 C n + a 2 a 3 q 3 a 1 a 2 a

Size: px

Start display at page:

Download "4 65 M = C p + f 1 C p f Sobel 1 C p f n C n 1 1 C n = B q n whn B n n w h q n F Q 13 7 F Q = a 1 C n + a 2 a 3 q 3 a 1 a 2 a"

恩蒋
7 years ago
Views:

1 40 4 Journal of South China University of Technology Vol 40 No Natural Science Edition April X * LIVE Video Quality Database TN919 8 doi /j issn X Naccari 2 Yang 3 1 Lee 4 Ichigaya PSNR 3 Yang * 2009BAF39B linxiangyu@ zju edu cn

2 4 65 M = C p + f 1 C p f Sobel 1 C p f n C n 1 1 C n = B q n whn B n n w h q n F Q 13 7 F Q = a 1 C n + a 2 a 3 q 3 a 1 a 2 a 3 M Q P M M 2 1 M Q P 4 M Q P D B Fig 1 Relationship between video subjective quality and quantization parameter

3 66 40 Fig 2 2 Screenshots of videos compressed with different rate control algorithms 3 DCT Fig 3 Bit allocation maps C RC C RC N B 2 4 A B S B 9 S I S I_AVG S Bi = A 15 i - B 0 i D x D y S I0 i = A 13 i - A 14 i S I1 i = A 14 i - A 15 i Δx Δy S I2 i = B 0 i - B 1 i 8 I M I M 1 N I I M S I3 i = B 1 i - B 2 i S { I_AVGi = S I0 i+ S I1 i + S I2 i + S I3 i I M = 1 槡 Δx2 + Δy 2 > T 1 槡 D x + D 4 y 4 0 i 0 15 T 1 3 C RC C RC = N I N B 100% 5 F B 1 3 Q S Q S Q S Q S = b 1 F B - b 2 F Q 7 b 1 b DCT F B = D B C RC 6 4 D B Fig 4 Distribution of pixels between adjacent macroblocks

4 4 67 S J J J i = { 0 S Bi- S I_AVGi S Bi > S I_AVGi 9 Sobel θ 0 Sobel a θ w S S 315 D BLOCK_H S D BLOCK_H = i = 0 J i wi w 10 { w 1 A 15 i B wi= 0 i w 2 A 15 i B 0 i 11 w 1 w 2 D BLOCK_H 6 D BLOCK_H_AVG Fig 6 Diagram of calculation of edge sharpness D BLOCK_V_AVG D BLOCK D 5 BLOCK D BLOCK = 1 2 D BLOCK_H_AVG + D BLOCK_V_AVG S S = 1 I j - I j = 1 j 5 I j 6b I 0 11 Sobel S D BLUR Fig 5 5 Comparison of clear and blur pictures Wang 12

5 JM Q D Q D 1 Q D 1 Q D = c 1 D BLOCK - c 2 D BLUK 14 Table 1 Weights and thresholds c 1 c 2 3 a T b T b 2 9 w c 1 1 w c t 10 s 20 a t 1 a 3 t 20 a t 1 a 3 c 1 c 2 2 LIVE LIVE 40 H 264 /AVC T Q M Q DMOS M P Q M Q M = Q D - Q S + d 1 d Spearman d 1 d 2 Q M β 1 - β 2 DMOS P = β 2 + LIVE Video Quality Database 1 + e - Q M 16 - β3 ( β ) LIVE 10 β 1 β 2 β 3 β Matlab 4 a a d 1 54 d SROCCPearson LCC 3 JM LIVE 1 LIVE Into skip 4 4 tree Stockholm Amazon Vidyo CABAC I 16

6 4 69 JM PSNR SSIM 15 SROCC LCC Table 2 2 The modification of encoder configuration I 30 6 I 8 7 B CAVLC SROCC LCC PSNR SSIM LIVE JM SROCC Fig 7 Scatter of experimental results based on different test LCC objects 3 3 LIVE 7 10 Table 3 Experiment results based on LIVE database 4 5 SROCC PSNR LCC SSIM MS-SSIM Speed-SSIM VSNR VQM V-VIF Spatial MOVIE Temporal MOVIE MOVIE GHz CPU 2 GB

7 70 40 Table 4 4 JM Experiment results based on JM encoded videos SROCC LCC PSNR SSIM Table 5 5Ichigaya A Kurozumi M Hara N et al A method of estimating coding PSNR using quantized DCT coefficients 5 J IEEE Transactions on Circuits System for Video Experiment results based on combined test videos Technology SROCC LCC PSNR SSIM Table 6 6 Computation complesity PSNR 79 8 SSIM /ms 9Seshadrinathan K Bovik A C Motion tuned spatio-temporal quality assessment of natural videos J IEEE Transactions on Image Processing Wang A Jiang G Wang X et al New no-reference blocking artifacts metric based on human visual system C International Conference on Wireless Communications and Signal Processing NanjingIEEE Wu S Lin W Xie S Blind blur assessment for visionbased applications J Journal of Visual Communication and Image Representation Wang X Tian B Liang C et al Blind image quality assessment for measuring image blur C Proceedings of International Congress on Image and Signal Processing SanyaIEEE Seshadrinathan K Soundararajan R Bovik A C et al Study of subjective and objective quality assessment of video J IEEE Transactions on Image Processing ITU-R BT Methodology for the subjective assessment of the quality of television pictures S 2Naccari M Tagliasacchi M Tubaro S No-reference video quality monitoring for H 264 /AVC coded video J IEEE Transactions on Multimedia Yang F Wan S Chang Y et al A novel objective no-reference metric for digital video quality assessment J IEEE Signal Processing Letters Lee S Jung K Sim D Real-time objective quality assessment based on coding parameters extracted from H 264 / AVC bitstream J IEEE Transactions on Consumer Electronic Yang F Wang S Xie Q et al No-reference quality assessment for networked video via primary analysis of bit stream J IEEE Transactions on Circuits System for Video Technology Seshadrinathan K Soundararajan R Bovik A C et al A subjective study to evaluate video quality assessment algorithms C Human Vision and Electronic Imaging XV San JoseSPIE2010 8ISO /IEC Advanced video coding S 14Sheikh H R Sabir M F Bovik A C A statistical evaluation of recent full reference image quality assessment algorithms J IEEE Transactions on Image Processing Wang Z Lu LBovik A C Video quality assessment based on structural distortion measurement C International Conference on Image Processing Rochester IEEE

8 78 40 Parallel Optimization of Three-Dimension Particle Simulation Based on Mixed MPI / OPENMP Programming Yan Xiao-yang Zhang Wei-wen Bu She-hui Deng Hui-fang School of Computer Science and EngineeringSouth China University of TechnologyGuangzhou GuangdongChina AbstractIn order to remedy the shortcomings of the parallel MPI program and the parallel OPENMP programa mixed parallel programming combining MPI and OPENMP is implemented Then the parallel processes of the MPI programthe OPENMP program and the mixed parallel MPI /OPENMP programming are respectively analyzed through experiments The results show that the mixed parallel MPI / OPENMP programming greatly reduces the communication trafficachieves an efficiency and a speedup rate higher than those of the parallel MPI programand a- voids the performance degradation of parallel MPI program due to the load imbalance caused by uneven particle distribution Thusmore nodes in the cluster can be used to perform a computation and the communication delay can be successfully decreased Moreover the mixed parallel programming significantly enlarges the simulation scale because it overcomes the dependency of parallel OPENMP program on the processing capacity and the storage space of a single computer Key wordsparallel MPI programparallel OPENMP program3dempmparallel computing 70 A Two-Domain No-Reference Video Quality Assessment Algorithm Lin Xiang-yu Chen Yao-wu Institute of Advanced Digital Technology and InstrumentationZhejiang UniversityHangzhou ZhejiangChina AbstractIn order to improve the accuracy and reduce the computational complexity of video quality assessmenta two-domain no-reference video quality assessment algorithm is proposed In this algorithm firstencoding information is extracted from the bitstream and video similarity is calculated according to the complexity of video contents in the compressed domain Thenby taking into consideration the characteristics of human visual systemthe video distortion is calculated by measuring the influence of blockiness and the blur artifacts in the pixel domain Thus the video quality is assessed according to the video similarity and the video distortion The proposed algorithm is finally tested on the LIVE Video Quality Database The results show that the proposed algorithmwhich possess a prediction accuracy and a monotonicity respectively up to and is better than the existing ten video quality assessment techniquesand that it meets the requirements of real-time application because it is of low computational complexity Key wordsvideo qualityno-reference assessment algorithmvideo similarityvideo distortionhuman visual system

34 www. cjig. cn wavelet transform 1 2 JPEG LIVE E s o = 1 T Σ log 2 C s o + 1 E T C s o Lu Wen contourlet C 0 7 N

34 www. cjig. cn wavelet transform 1 2 JPEG LIVE E s o = 1 T Σ log 2 C s o + 1 E T C s o Lu Wen contourlet C 0 7 N 17 1 2012 1 Journal of Image and Graphics Vol. 17No. 1 Jan. 2012 TP391. 4 A 1006-8961201201-0033-07. NSS J. 2012 171 33-39 NSS 214122 NSS No-reference image quality assessment based on natural scene statistics