Curriculum Vitaes
  1. Researcher Search Results
  2. Kenji Sugiyama

Kenji Sugiyama

  (杉山 賢二)

Profile Information

Affiliation
Professor, Faculty of Science and Technology Department of Science and Technology , Seikei University
Degree
博士（工学）(東京理科大学)
J-GLOBAL ID
201501074355666890
researchmap Member ID
B000245394
成蹊大学理工学部教授

Research Interests

 3

Research Areas

 2

Research History

 2

Education

 3

Committee Memberships

 10
More

Awards

 1

Papers

 33
  • Takashi Yamazoe, Jinyu Tang, Gin Inoue, Kenji Sugiyama
    IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Sep, 2023  
  • Go Kasahara, Takashi Yamazoe, Kenji Sugiyama
    Digest of Technical Papers - IEEE International Conference on Consumer Electronics, 2020-January, Jan, 2020  Peer-reviewedLast author
    © 2020 IEEE. It is common in color imaging to use a single-sensor system with a color filter array (CFA), especially a Bayer CFA. Recently, color filter methods using the white component have been studied to improve both resolution and sensitivity. However, de-mosaicking performance is improved using a correlation of color planes, especially for a Bayer CFA. We propose a new imaging system with a pastel-color CFA. In this system, the color plane reconstruction is realized with an abstraction of the white-like component from RAW images. To improve performance, we discuss the use of the highfrequency component. However, the effect of sensor noise on the imaging performance has not been closely investigated in such imaging systems. In this paper, we evaluate these effects on various imaging systems. In our experiment, the optical color filter and sensor noise are simulated using computer calculations. With these results, it is recognized that the proposed method is less affected by sensor noise.
  • Yoshikazu Honma, Kenji Sugiyama
    IEEE Proceedings of International Conference on Consumer Electronics (ICCE2019), Jan, 2019  Peer-reviewed
  • Yoshikazu Honma, Naoyuki Awano, Kenji Sugiyama
    2018 IEEE International Conference on Consumer Electronics, ICCE 2018, 2018- 1-4, Mar 26, 2018  Peer-reviewed
    Currently, Y, Cb, Cr 4:2:0 system is widely used as the color format for the video system. We have proposed the frame sequential system as an alternative method which has one color plane in a frame. In this system, the conversion to usual RGB requires a motion compensated color interpolation. However, it is not easy in the case of high saturation color images, because of its difference between color planes. To avoid this problem, we propose the new color format which has dependent colors. The primary colors (RGB) are changed to pastel colors. The conversion from RGB to pastel colors is realized by the color matrix. Re-conversion is also possible by using the inverse matrix. To address the conversion to RGB, basic block matching is used in motion estimation. To avoid incorrect motion vectors, the antierror mechanism are strengthened. Finally, the performance of the pastel colored frame sequential and 4:2:0 are compared. From experimental results, the converted RGB pictures have higher PSNR than 4:2:0.
  • Ayumu Ohike, Naoyuki Awano, Kenji Sugiyama
    2016 IEEE INTERNATIONAL CONFERENCE ON CONSUMER ELECTRONICS (ICCE), 526-529, 2016  Peer-reviewed
    Very high-resolution video systems, such as 4K (3840x2160), enable a very close viewing distance that is almost the same as the picture height. This technology enables high-reality systems in homes to be realized. However, the very short distance causes significantly different views across the picture, and uniform processing for the picture may not always be the best choice. Here, we calculate the actual view size of a pixel, which depends on the pixel position in close viewing. We proposed the perceptual coding methods using pre/post-processing for the picture to be coded. It is a quantization control by a contrast change. However, the control of quantization should be decided based on the perceptual performance between pixel size and quantization. Therefore, we carried out subjective assessment about this. The bit rate reduction is checked for each method. The 4K resolution processing is realized by four patterns of HDTV picture coding with AVC. As the results of these experiments, it is recognized that the bit rate reduction 15 percent by the optimal coding.
More

Misc.

 42
  • 117(145) 1-4, Jul 21, 2017  
  • 117(35) 21-26, May 19, 2017  
  • 41(6) 9-12, Feb, 2017  
  • 28(1) 17-23, Jan, 2017  
  • 116(369) 19-24, Dec 16, 2016  
  • 40(35) 17-20, Oct, 2016  
  • 116(68) 7-12, May 27, 2016  
  • SUGIYAMA Kenji, AWANO Naoyuki, SASAKI Ryota
    114(485) 103-106, Mar 3, 2015  
    Realizing of the super high definition video (4K/8K) require the format conversion between different picture formats and the improvement of image quality of the converted image. To suppress the coding cost, pixel number reduction to be coded is discussed. In these system, picture quality of up converted picture, such as HDTV to 4K, is most important. In this report, the subjective assessment for the scaled images processed by different scaling factor and filter order are studied. The enhance processing to improvement sharpness with increase the high frequency component in residual band. Frequency characteristic of filter is discussed. The subjective assessments are executed with double stimulation. The results shows that the enhancement gives better MOS. This means that the lack of high frequency component can be compensated by enhancement of relatively low frequency component.
  • TAKANO Hirofumi, INOMATA Mitsumi, AWANO Naoyuki, SUGIYAMA Kenji
    114(485) 183-186, Mar 3, 2015  
    To realize the subjective assessment for high dynamic range (HDR) images, we have used professional use organic electroluminescence (EL) display which has enough contrast performance. In this, the pseudo 16 bit HDR images were converted to BT.709 standard for the assessment. However, there may be problem about displayed image, therefore, re-examination were done. We also have tried to use a normal LCD with tone conversion which has knee point. The correlation between the result of the organic EL assessment and LCD assessment reached to 0.96. About the objective assessment, We propose the weighted SNRs of the HDR images using the two dimensional frequency characteristics of the contrast sensitivity. Because they require FFT, the simpler method using one contrast sensitivity value based on ASNR are discussed. Finally, the correlation between the results of EL assessment and CASNR are almost the same as WASNR.
  • SUGIYAMA Kenji, AWANO Naoyuki, OIKE Ayumu, KUSAJIMA Takuya
    ITE Technical Report, 39(6) 25-28, Feb, 2015  
    Very high resolution video system such as 4K/8K enables super close view, and, such very short distance causes significant different view in a picture. On the other hand, the human gaze mostly distribute to the center of the picture, and this is more significant in the large screen. Therefore, uniform coding for a picture may not be always best. In this report, the reduction of the number of pixels to be coded using the geometric image conversion is discussed on the viewpoint of processing power and coding efficiency. The basic method with two dimensional conversion based on the actual view and the horizontal vertical independent methods are compared. Furthermore, the quantization control using the contrast conversion to reduce the residual information in the independent method is discussed. As the results of experiments, the possibility of maximum 35 percent bit rate reduction is recognized with 50% processing power.
  • OHIKE Ayumu, AWANO Naoyuki, SUGIYAMA Kenji
    114(386) 29-32, Dec 20, 2014  
    Very high-resolution video systems, such as 4K, enable a very close viewing distance. This technology enables high-reality systems in homes to be realized. However, the very short distance causes significantly different views across the picture, and uniform processing for the picture may not always be the best choice. Here, we calculate the actual view size of a pixel, which depends on the pixel position in close viewing. We have proposed the two kinds of perceptual coding methods. First one is a pre-filtering which depend on the size of a pixel. Another one is a quantization control by a contrast change. However, the control of quantization should be decided based on the perceptual performance between pixel size and quantization. Therefore, we carried out subjective assessment about this. The bit rate reduction is checked for each methods. As the results of this experiments, it is recognized that up to 15% of bit rate reduction is achievable.
  • SUGIYAMA Kenji, HONDO Yuta, SUZUKI Ai, AWANO Naoyuki
    114(35) 13-16, May 16, 2014  
    About the subjective assessment for High Dynamic Range images, usual assessment is enough to evaluate tone conversion to Low DR images. However, a special assessment of HDR images is required to evaluate the coding processing. To realize this, we have tried to use a normal LCD and tone conversion with knee point characteristics. This method is useful, however, not enough, and, more reasonable device and environment for HDR images are required. This time, we use professional use organic electroluminescence display which has enough contrast performance. Tested pseudo 16 bit HDR images are converted to BT.709 standard. As the results of experiments, it was recognized that an assessment of HDR images are possible without tone conversion at dark room viewing. However, reliable assessment is not easy on the very dark image which has almost 0.4% linear average brightness.
  • 粟野直之, 杉山賢二
    第76回全国大会講演論文集, 2014(1) 59-60, Mar 11, 2014  
    可視化を必要とする情報の一つに，曲線や曲面の曲がり具合を端的に表す形状特徴量の曲率がある．曲率を可視化する際には，一般に曲率の値と対応させたカラーパレットで可視化する．しかし，現在広く使用されているカラーパレットでは曲率と色の組み合わせが直観的でない可能性がある．本研究では，曲率と色の組み合わせが最適となる組み合わせを実験により評価し，決定する．ここで，曲率は角度を基に算出されることに着目し，角と色との関係を主観評価実験により評価する．実験では色から連想する角と，角から連想する色を評価し，角に対する色を決定する．そして，従来のカラーパレットを使用した配色等との比較実験を行う．
  • OKADA Yuki, AWANO Naoyuki, SUGIYAMA Kenji
    Technical report of IEICE. Multimedia and virtual environment, 113(470) 129-132, Mar 6, 2014  
    We have proposed the sequential color system for use as a video format. To realize this system, the conversion to whole RGB picture with the interpolation of missing color plane using motion compensation is required. In the motion estimation of sequential color system that three color planes are changed sequentially by a frame, since each same color has a frame distance of three, motion estimation in different color plane is required to adapt for area with fast and complicated motion. On the other hand, since the color is different between one frame and two frame in different color plane, one frame distance is not necessarily appropriate. Therefore, we propose how to choose motion vectors obtained from three kinds of color plane relationships including same color plane. As results of experiments, it is recognized that PSNR and subjective picture quality of conversion picture is improved.
  • KUSAJIMA Takuya, MATSUMOTO Satoshi, AWANO Naoyuki, SUGIYAMA Kenji
    ITE Technical Report, 38(8) 5-8, Feb, 2014  
    In re-size (scaling) of a picture, quality up processing which are called super resolution have been studied. We have discussed a method of directional adaptive processing to reduce jaggy and ringing. On the other hand, geometric image conversion such as affine transform is one of re-size processing which have different magnification scale in a pixel. However, some super resolution methods are not suitable to this processing. At first, we discuss a hybrid re-sampling which uses a high order filter and a linear interpolation to balance the spatial frequency and flexibility. Next, pre-filtering is discussed to reduce alias component in size deduction. Finally, we apply a directional adaptive processing for the geometric image conversion to reduce jaggy. As results of experiments, it was recognized that the proposed method gives better PSNR and subjective picture quality.
  • TAKANO Hirofumi, AWANO Naoyuki, SUGIYAMA Kenji
    ITE Technical Report, 38(8) 13-16, Feb, 2014  
    A pixel (sample) of high dynamic range (HDR) image require 16 bit quantization without gamma correction. Standard formats for the broadcast are 10 bit with gamma correction, and not suitable for file storage system because of non byte alignment. To realize high versatility and utility for HDR images, we proposed the range coding with 4 byte for 4 pixels. This data is easily changeable to 16bit linear quantized pixels. In this report, we discuss the quality improvement method using adaptive range coding based on histogram of pixel value. Narrower range is applied to the dense level with keeping range number of 16. With In the case of LPSNR (local peak signal to noise ratio) which is suitable for the evaluation of HDR images, almost 1 dB SNR gain is achieved for dark images, and it is recognized that performance of proposed method is almost the same of 10bit with gamma correction.
  • OIKE Ayumu, KUSAJIMA Takuya, AWANO Naoyuki, SUGIYAMA Kenji
    ITE Technical Report, 38(8) 9-12, Feb, 2014  
    Very high resolution video system such as 4k (4096x2160) enables super close view that distance is almost same as picture height. However, very short distance causes significant different view in a picture. Therefore, uniform processing for a picture may not be always best. At first, we calculate the actual view size of a pixel depends on the pixel position in the case of a picture height distance. the coding performance is checked for the weighted picture in frequency band limitation, contrast conversion and geometric image conversion the coding performance is checked for the weighted picture in frequency band limitation, contrast conversion and geometric image conversion. 4k resolution processing is realized by 4 pattern of HDTV picture coding with H.264 and HEVC. As the results of experiments, the possibility of 10 to 25 percent bit rate reduction is recognized.
  • SUGIYAMA Kenji, SAGARA Naoya, NAKAMURA Koji, FURUTA Syou
    113(45) 1-6, May 17, 2013  
    Very high resolution video system such as 4k (4096x2160) enables super close view that distance is almost same as picture height. However, very short distance causes significant different view in a picture. Therefore, uniform processing for a picture may not be always best. At first, we calculate the actual view size of a pixel depends on the pixel position in the case of a picture height distance. To estimate the effect, the picture quality assessments were done with the pictures which include weighted basic degradations such as gradient and white noise. As the results, it is recognized that the subjective quality of weighted pictures are better than uniform one. Furthermore, the coding performance is checked for the weighted picture in frequency band limitation and contrast conversion. As the results of experiments, the possibility of 10 to 30 percent bit rate reduction is recognized.
  • MATSUMOTO Satoshi, SAGARA Naoya, SUGIYAMA Kenji
    112(161) 31-34, Jul 20, 2012  
    In re-size (scaling) of a picture, a directional adaptive processing which realizes an improvement of subjective picture quality with a reduction of jaggy is discussed. In this processing, the clustering to change the processing by area type and the directional adaptive processing at edge area are applied. Although the authors had used the operators to detect 12 directions, two times magnification was realized to prevent the complexity of the reference pixel. In this paper, we study the 12 directional adaptive interpolation with four times magnification. In the expanding with high magnification, the direction adaptation is more effective, because jaggies are remarkable. The pixels used in interpolation are the original pixel and interpolated pixel which were made by one dimensional 8 tap filter. The pixels which are not nearest are used for the directions close to 0 or 90 degree. Nearest pixels are used for 45 or close degree. PSNR and enlarged picture show good results.
  • SUGIYAMA Kenji, SAGARA Naoya, HONDO Yuta
    112(40) 13-18, May 11, 2012  
    Recent all standard video formats are based on BT.601 which was recommended in early 80'. This standard was aimed to keep all of analogue video signal with the gamma conversion (compensation) and 8 or 10 bit quantization includes margin. HDTV standards which were recommended in mid of 90' were changed to square pixel, however, the structure of pixel value were almost kept. On the other hand, PC application of video was dramatically changed, and the correspondence to HDR (High Dynamic Range) image becomes important. Therefore, it is doubtable that the digital video formats based on the analogue extension is best. To realize high versatility and utility of video, we discuss the range coding with plural pixels combination. Further more, to extend tone, the quantization method with offset threshold is proposed. To test the picture quality, pseudo 16 bit HDR images made from 10 bit HDTV are used.
  • SUGIYAMA Kenji, SAGARA Naoya, HONDO Yuta
    111(459) 7-10, Feb 24, 2012  
    To coding the high dynamic range (HDR) images, the quantization with the conversion based on the tone mapping is usually discussed. This method is suitable to view with normal (low) dynamic range display devices, but not suitable to the high fidelity transmission of HDR images. To realize high versatility and utility coding for the HDR images, we discuss the range coding with plural pixels combination. About the evaluation of coding for the HDR images, new type of SNR to use the independent pixel value for as a signal instead of the peak value is proposed. Tested pseudo 16 bit HDR images are made from 10 bit HDTV standard video with the conversion processing to evaluate many kind of images. As the results of experiments, simple range coding is poor than gamma conversion, however, the proposed coding, especially 4 pixel combine method gives best performance.
  • OKADA Yuki, HASEGAWA Toshio, SAGARA Naoya, SUGIYAMA Kenji
    ITE Technical Report, 36(10) 9-12, Feb 17, 2012  
    We have proposed a sequential color system for use as a video format. In this system, three color planes are used sequentially, a conversion to whole RGB picture with the interpolation of missing color plane is required. However, conventional motion estimation is not useful because of the changing of color plane by a frame. To respond this, we have proposed the motion estimation method using the contour picture. However, the difference between the contour of different color is not small in high color picture. In this paper, we propose the normalizing processing for contour picture by area level. The error reduction processing for motion vector is also applied. As results of experiments, it is recognized that PSNR and subjective picture quality of motion compensated picture is improved significantly.
  • SUZUKI Kazuki, SAGARA Naoya, SUGIYAMA Kenji
    111(318) 7-10, Nov 25, 2011  
    The sequential color is proposed for use as a new video format. On the other hand, RGB direct coding is noticed because of the problem of luminance and chrominance system. We consider RGB direct coding methods of picture quality control by luminance and chrominance. Then, we proposed the subjective quality improvement method by enhancing the PSNR of luminance component using quantization error diffusion between the color planes. However, PSNR of the chrominance components is decreased, and too large coefficient causes the degradation of subjective quality. As results of subjective assessment, 0.5 MOS gain is achieved with 0.5 to 1.0 coefficients of error diffusion.
  • Hasegawa Kazuhiro, Sagara Naoya, Sugiyama Kenji
    Proceedings of the Society Conference of IEICE, 2011 "S-59"-"S-60", Aug 30, 2011  
  • SUZUKI Takayuki, SAGARA Naoya, SUGIYAMA Kenji
    IEICE technical report, 111(192) 23-26, Aug 26, 2011  
    In the many kinds of video system, it is desired that the picture quality is controlled to provide the better picture. For this purpose, an objective picture quality estimation method without the reference picture is required. However, the objective estimation is not easy in the display, because of lack of the bit-stream information. To respond this, we discuss the non-referenced general estimation method using the reported estimation methods of the block artifact, the mosquito noise and the temporal degradation, further, a spatial blur estimation using DCT. This method works to give the absolute results which match to the subjective picture quality using reconstructed picture only. The results of general estimation are given as the estimated value for the subjective quality of reconstructed picture. In the comparison, high correlation were measured for each sequences, however, overall performance is not enough.
  • SUGIYAMA Kenji
    IEICE technical report, 111(35) 33-38, May 6, 2011  
    Currently, the widely used video formats are 4:2:0 or 4:2:2, however, the picture quality is degraded; in particular with interlaced 4:2:0, the degradation of the luminance signal is significant with high saturation color. In addition, as a format over the HDTV, Super Hi-Vision and 4k by 2k (2160p) are discussed. However, it seems that these formats have very high spec as consumer application. On the other hand, I have proposed the new video format called the sequential color system, in which three color planes appear sequentially. Sequential color retains the picture quality of 4:4:4, as long as the racked color is carefully interpolated. Therefore I reexamine the good point of the sequential color method, and propose the video formats for consumer application using it. And, I compare the performance of each format and examine the feasibility of the format using the sequential color system.
  • SUZUKI Kazuki, SAGARA Naoya, SUGIYAMA Kenji
    Proceedings of the ... ITE annual convention, 7-10-1-"7-10-2", 2011  
    Sequential color has only a plane in a frame and each planes are switched for each frame. To realize this system coding efficiency is important. Therefore, the coding efficiency of luminance/chrominance system and RGB system are compare at Intra coding.
  • 杉山賢二
    電子情報通信学会 イメージメディアクウォリティ研究会, 2010 25-30, 2010  
  • 杉山賢二
    電子情報通信学会 イメージメディアクウォリティ研究会, 2010 25-30, 2010  
  • 杉山賢二
    電子情報通信学会 イメージメディアクウォリティ研究会, 2010 25-30, 2010  
  • 杉山 賢二
    画像ラボ, 19(11) 53-57, Nov, 2008  
  • YAMAMOTO Yuya
    ITE technical report, 32(11) 17-20, Feb, 2008  
  • The journal of the Faculty of Science and Technology, Seikei University, 44(2) 45-52, Dec, 2007  
  • ITE technical report, 31(13) 1-4, Feb, 2007  
  • ITE technical report, 31(13) 5-8, Feb, 2007  
  • SUGIYAMA Kenji, YAMADA Yoshiyuki
    ITE technical report, 30(16) 25-28, Feb 24, 2006  
  • ANDO Akira, HANGAI Seiichiro, SUGIYAMA Kenji
    ITE technical report, 29(15) 17-20, Feb 25, 2005  
  • SUGIYAMA Kenji, OBARA Gou, HANGAI Seiichiro
    IEICE technical report. Communication systems, 100(500) 121-126, Dec 8, 2000  
    Abstract : In digital broadcast, many kind of picture format are used. However, interlaced scanning TV receiver is still used in most homes. We have thus developed a broadcast system that uses common bit stream for the 480i and the 480p formats as a source picture and the 480i format for the receiver. We take new coding for the 480i pictures which include a conversion from interlaced scanning to progressive scanning of the P and I-picture only. We discuss a scanning conversion and use motion compensated adaptive interpolation. Experimental results showed that the proposed coding scheme performs better than interlaced scanning coding in MPEG-2.
  • 杉山 賢二, 藤原 光章, 横内 健太郎
    テレビジョン学会誌, 49(9) 1209-1209, Sep 20, 1995  
  • SUGIYAMA Kenji, NAKAMURA Hiroya
    IEICE technical report. Image engineering, 95(180) 63-68, Jul 24, 1995  
    In the motion compensation, small block size is required for a good motion estimation. Small block size often occur incorrect motion vector, because of a little samples for a block matching. We have investigated a method considering surrounding blocks. In this method, motion vector is not determined by block matching directory, but determined by the result of adding about surrounding blocks. Adding factor is adapted to a similarity of high priority motion vectors.
  • SUGIYAMA Kenji, FUJIWARA Mitsuaki, YOKOUCHI Kentaro
    ITEJ Technical Report, 19(19) 7-12, Mar 17, 1995  
    In the compact-sized optical disc for video application, 135 minute or more recording time is required. However, current recording and coding techniques could not satisfying this purpose. On the other hand, variable bit rate coding gives ideal picture quality. We applied this scheme to the disc, and evaluated encoding, recording, playback and decoding. In this report, rate control method is discussed mainly. It gives constant visual quality and the data amount fitting in the disc capacity.
  • 杉山 賢二
    テレビジョン学会誌, 44(3) 358-358, Mar 20, 1990  

Books and Other Publications

 4

Teaching Experience

 3

Professional Memberships

 3

Research Projects

 2

Industrial Property Rights

 14
More
To the list screen