Kyushu University Academic Staff Educational and Research Activities Database
List of Papers
Kenji Hara Last modified dateļ¼š2023.11.28

Professor / Department of Visual Communication Design / Department of Media Design / Faculty of Design


Papers
1. Zilin Xiao, Kohei Inoue, Kenji Hara, Naoki Ono, Toru Hiraoka, Generating Variable-Size Pixel Art Based on Row/Column Clustering, ICIC Express Letters, 10.24507/icicel.16.09.965, 16, 9, 965-972, 2022.09.
2. Kohei Inoue, Naoki Ono, Kenji Hara, Local Contrast-Based Pixel Ordering for Exact Histogram Specification, Journal of Imaging, 10.3390/jimaging8090247, 8, 9, 2022.09.
3. Qi Zheng, Kohei Inoue, Naoki Ono, Kenji Hara, A Hough Transform for Detecting Facial Symmetry Axis, Journal of the Institute of Industrial Applications Engineers, 10, 13, 53-59, Vol. 10 No. 3, pp. 53--59, 2022.07.
4. Ran Peng, Kohei Inoue, Kenji Hara, Generating Face Images Based on Golden Ratios for Facial Beauty, Journal of the Institute of Industrial Applications Engineers, 10, 1, 11-18, Vol. 10 No. 1, pp. 11--18, 2022.01.
5. Shuoyan Zhang, Kohei Inoue, Kenji Hara, A convergence proof for local mode filtering, ICT Express, 7, 4, 445-448, Volume 7, Issue 4, December 2021, Pages 445-448, 2021.12.
6. Takayuki HATTORI, Kohei Inoue, Kenji Hara, Rolling Guidance Filter as a Clustering Algorithm, IEICE TRANSACTIONS on Information and Systems, E104-D, 10, 1576-1579, Vol.E104-D, No.10, pp.1576--1579, 2021.10.
7. Shanqian Sun, Kohei Inoue, Kenji Hara, Adaptive Combination of Additive and Multiplicative Algorithms for Color Image Enhancement, Journal of the Institute of Industrial Applications Engineers, 9, 2, 52-59, 2021.04.
8. Wenyi Cui, Kohei Inoue, Kenji Hara, An Ambigramic Image File Format, Global Journal of Computer Science and Technology (GJCST), 21, 1, 1-5, F, Graphics and Vision, 21, 1, 1-5, F, 2021.03.
9. Kohei Inoue, Kenji Hara, An equivalence between log-sum-exp approximation and entropy regularization in K-means clustering, Nonlinear Theory and Its Applications, IEICE (NOLTA), 10.1587/nolta.11.446, 11, 4, 446-453, 2020.10.
10. Wei Chen, Kohei Inoue, Kenji Hara, Adaptive Aggregated Histogram Equalization for Color Image Enhancement without Gamut Problem, Journal of the Institute of Industrial Applications Engineers, 10.12792/jiiae.8.56, 8, 2, 56-62, 2020.04.
11. Kaiming Wu, Kohei Inoue, Kenji Hara, Neugebauer Models for Color Error Diffusion Halftoning, Journal of Imaging, 10.3390/jimaging6040023, 6, 4, 2020.04.
12. Daiki Okazaki, Kenji Hara, Kohei Inoue, Kiichi Urahama, Optimum Parameter Estimation of Tone Mapping Operators by Natural Image Statistics, Journal of the Institute of Industrial Applications Engineers, 7, 3, 106-109, 2019.10.
13. Ruixi Dong, Kohei Inoue, Kenji Hara, Kiichi Urahama, Two-in-One Image Steganography Using Error Diffusion, Journal of the Institute of Industrial Applications Engineers, 7, 2, 42-50, 2019.04.
14. An shui Yu, Kenji Hara, Kohei Inoue, Kiichi Urahama, Foreground enlargement of spherical images using a spring model, IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, 10.1587/transfun.E102.A.486, 2, 486-489, 2019.02.
15. Atsushi Morinaga, Kenji Hara*, Kohei Inoue, Kiichi Urahama, Classification between Natural and Graphics Images Based on Generalized Gaussian Distributions, Information Processing Letters (IPL), https://doi.org/10.1016/j.ipl.2018.05.010, 138, 31-34, 2018.10.
16. Hengjun Yu, Kohei Inoue, Hara Kenji, Kiichi Urahama, Saturation improvement in hue-preserving color image enhancement without gamut problem, ICT Express, 10.1016/j.icte.2017.07.003, 4, 3, 134-137, 2018.09, This study proposes a method to improve the saturation of colors during hue-preserving color image enhancement without the gamut problem. First, the color of each pixel in an input image is projected onto one of three bisecting planes or bisectors of an RGB color cube to increase saturation. Then, that projected color is transformed into a target color with a designated luminance and the same hue as the original color. Experimental results show that the proposed method can enhance color images better than can recent hue-preserving color image enhancement methods..
17. Hengjun Yu, Kohei Inoue, Kenji Hara, kiichi urahama, Saturation-Improved Hue-Preserving Gamut Mapping by Weighted Histogram Specification, Journal of the Institute of Industrial Applications Engineers, 6, 3, 112-118, 2018.07.
18. Hengjun Yu, Kohei Inoue, Kenji Hara, Kiichi Urahama, A Robust K-Means for Document Clustering, Journal of the Institute of Industrial Applications Engineers, 6, 2, 60-65, 2018.04.
19. Mai Shikita, Kohei Inoue, Hara Kenji, Kiichi Urahama, Continuity-aware estimation of spectral reflectance from tristimulus values, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.72.J105, 72, 8, J105-J108, 2018.01.
20. Mai Shikita, Kohei Inoue, Hara Kenji, Kiichi Urahama, Estimating reflectance spectra from tristimulus values under different illuminations, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.72.J86, 72, 5, K86-J88, 2018.01.
21. An-shui YU, Naoki Ono, Kohei Inoue, Hara Kenji, Kiichi Urahama, Backward Filtering of Images by Iterating Forward Filters, Journal of the Institute of Industrial Applications Engineers, 5, 4, 197-203, 2017.10, Given a filtering operation for images, we consider a problem for inverting it. If the algorithm of the original,i.e. forward filter is explicitly expressed, its inversion is tractable. However, if its algorithm is unknown exactly, it is hard to be reversed. We present, in this paper, a simple solution for this problem only by iterating the forward filter. This method can be used for a wide class of image filters including ones built in photo-retouch softwares. Examples are shown for smoothing, sharpening, halftoning and resizing filters..
22. Kohei Inoue, Hara Kenji, Kiichi Urahama, RGB Color Cube-Based Histogram Specification for Hue-Preserving Color Image Enhancement, Journal of Imaging, 3, 3, 2017.09.
23. Kohei Inoue, Hara Kenji, Kiichi Urahama, RGB Color Cube-Based Histogram Specification for Hue-Preserving Color Image Enhancement, Journal of Imaging, 10.3390/jimaging3030024, 3, 3, 2017.09.
24. Kohei Inoue, Hara Kenji, Kiichi Urahama, RGB color cube-based histogram specification for hue-preserving color image enhancement, Journal of Imaging, 10.3390/jimaging3030024, 3, 3, 2017.09, A large number of color image enhancementmethods are based on themethods for grayscale image enhancement in which the main interest is contrast enhancement. However, since colors usually have three attributes, including hue, saturation and intensity of more than only one attribute of grayscale values, the naive application of themethods for grayscale images to color images often results in unsatisfactory consequences. Conventional hue-preserving color image enhancement methods utilize histogram equalization (HE) for enhancing the contrast. However, they cannot always enhance the saturation simultaneously. In this paper, we propose a histogram specification (HS) method for enhancing the saturation in hue-preserving color image enhancement. The proposed method computes the target histogram for HS on the basis of the geometry of RGB (rad, green and blue) color space, whose shape is a cube with a unit side length. Therefore, the proposed method includes no parameters to be set by users. Experimental results show that the proposed method achieves higher color saturation than recent parameter-free methods for hue-preserving color image enhancement. As a result, the proposed method can be used for an alternative method of HE in hue-preserving color image enhancement..
25. An-shui Yu, Naoki Ono, Kohei Inoue, Kenji Hara, kiichi urahama, Isoluminant Color Halftoning for Generating Pointillistic Tile Images and Photomosaics, Journal of the Institute of Industrial Applications Engineers, 5, 3, 104-110, 2017.07.
26. An-shui YU, Naoki Ono, Kohei Inoue, Hara Kenji, Kiichi Urahama, Isoluminant Color Halftoning for Generating Pointillistic Tile Images and Photomosaics, Journal of the Institute of Industrial Applications Engineers, 5, 3, 104-110, 2017.07.
27. Kohei Inoue, Hara Kenji, Kiichi Urahama, Spectral Reflectance Estimation and Color Reproduction Based on Sparse Neugebauer Model,, Advances in Science, Technology and Engineering Systems Journal, 2, 3, 958-966, 2017.06.
28. Kohei Inoue, Kenji Hara, kiichi urahama, Hengjun Yu, Histogram Interpolation Methods for Image Contrast Enhancement, Journal of the Institute of Industrial Applications Engineers, vol. 5, no. 1, 7-14, 2017.01.
29. Zixu Zhang, Kohei Inoue, Hara Kenji, Kiichi Urahama, Color image enhancement based on neugebauer model, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.71.J155, 71, 4, J155-J158, 2017.01.
30. An-shui YU, Hara Kenji, Kohei Inoue, Kiichi Urahama, Foreground Enlargement onSpherical Images Based on Analytical Mechanics, International Workshop on Advanced Image Technology, 5, 3, 2017.
31. Kohei Inoue, Hara Kenji, Kiichi Urahama, Spectral Reflectance Estimation and Color Reproduction Based on Sparse Neugebauer Model, Advances in Science, Technology and Engineering Systems Journal, 10.25046/aj0203121, 2, 3, 958-966, 2017.
32. Hengjun Yu, Kohei Inoue, Kenji Hara, kiichi urahama, Color Error Diffusion Based on Neugebauer Model, IEICE Trans. Fund., vol. E99-A, 3, 1758-1761, 2016.09.
33. Hengjun Yu, Kohei Inoue, Hara Kenji, Kiichi Urahama, Color error diffusion based on neugebauer model, IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, 10.1587/transfun.E99.A.1758, E99A, 9, 1758-1761, 2016.09, In this paper, we propose a method for color error diffusion based on the Neugebauer model for color halftone printing. The Neugebauer model expresses an arbitrary color as a trilinear interpolation of basic colors. The proposed method quantizes the color of each pixel to a basic color which minimizes an accumulated quantization error, and the quantization error is diffused to the ratios of basic colors in subsequent pixels. Experimental results show that the proposed method outperforms conventional color error diffusion methods including separable method in terms of eye model-based mean squared error..
34. Kohei Inoue, Kenji Hara, kiichi urahama, Applying Color Schemes to Color Images with Unilateral Filter, Journal of the Institute of Industrial Applications Engineers, vol. 4, no. 3, 118-121, 2016.07.
35. Kohei Inoue, Hengjun Yu, Hara Kenji, Kiichi Urahama, Saturation-enhancing multi-exposure image fusion, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.70.J185, 70, 8, J185-J187, 2016.01.
36. An-shui YU, Hara Kenji, Kohei Inoue, Kiichi Urahama, A Corner Detection Approach for Fisheye Images, International Workshop on Advanced Image Technology, 5, 3, 2016.
37. Hengjun Yu, Kohei Inoue, Kenji Hara, kiichi urahama, Image Contrast Enhancement by Stepwise Histogram Specification, Journal of the Institute of Industrial Applications Engineers, vol. 3, no. 4, 161-166, 2015.10.
38. Kohei Inoue, Kenji Hara, kiichi urahama, On Hue-Preserving Saturation Enhancement in Color Image Enhancement, IEICE Trans. Fund., E98-A, 3, 927-931, 2015.03.
39. Hara Kenji, Kohei Inoue, Kiichi Urahama, Gradient operators for feature extraction from omnidirectional panoramic images, Pattern Recognition Letters, 10.1016/j.patrec.2014.12.010, 54, 89-96, 2015.03, We propose an extension of the Laplacian operator and the class of differential corner detectors for omnidirectional panoramic images covering 360° views. It solves two problems of nonuniformity of spatial resolution and spherical polar coordinates singularity. For the first problem, we derive a Laplacian operator and a class of differential corner detectors that can handle spatial inhomogeneities in omnidirectional panoramic images using differential schemes. To solve the second problem, we use an overset grid system on the sphere called the Yin-Yang grid. This is realized based on the property that each of the modified Laplacian operator and corner detector is invariant with respect to coordinate transformation between the two spherical polar coordinate systems associated with the Yin-Yang grid system..
40. Kohei Inoue, Hara Kenji, Kiichi Urahama, On hue-preserving saturation enhancement in color image enhancement, IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, 10.1587/transfun.E98.A.927, E98A, 3, 927-931, 2015.03, Recently, hue-preserving color image enhancement methods have been proposed by several researchers. However, the theoretical comparison of the performance of their methods has not been conducted yet. In this paper, we propose a hue-preserving saturation maximization method, and show a relationship of the saturation of enhanced colors by related methods. We also demonstrate the correctness of the relationship experimentally..
41. Kenji Hara, Kohei Inoue, kiichi urahama, Gradient Operators for Feature Extraction from Omnidirectional Panoramic Images, Pattern Recognition Letters (PRL), vol. 54, pp. 89-96, 2015.01.
42. An-shui YU, Hara Kenji, Kohei Inoue, Kiichi Urahama, Spatially-Variable Laplacian Edge Detector for Fisheye Images, International Workshop on Advanced Image Technology, 5, 3, 2015.
43. Kenji Hara, Kohei Inoue, kiichi urahama, A Differentiable Approximation Approach to Contrast-Aware Image Fusion, IEEE Signal Processing Letters (SPL), 10.1109/LSP.2014.2314647, vol.21, no.6, pp.742-745, 2014.06.
44. Kohei Inoue, Katsuya Suenaga, Hara Kenji, Kiichi Urahama, Miniature faking using smoothing and sharpening, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.68.J108, 68, 2, 2014.01.
45. Kohei Inoue, Hara Kenji, Kiichi Urahama, Edge-Preserving in centroidal voronoi tessellation-based halftoning, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.67.J410, 67, 11, 2013.11, Conventional centroidal Voronoi tessellation (CVT) in an image plane divides the plane into several Voronoi regions. Generally, the boundaries of Voronoi regions do not coincide with the edges of the image. Therefore, image segmentation based on CVT often disturbs the edges. We developed an edge-preserving CVT technique that reduces the disturbance around the edges of an image and the Voronoi boundaries and applied it to non-photorealistic halftoning..
46. Size-Adaptive Hepatocellular Carcinoma Detection from 3-D CT Images Based on the Level Set Method.
47. Kohei Inoue, Hara Kenji, Kiichi Urahama, Gradient norm-based histogram equalization, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.67.J296, 67, 8, 2013.09, We propose a method for enhancing the contrast of grayscale images by equalizing their weighted histograms on the basis of the gradient norm, and we quantitatively evaluate the effectiveness of the proposed method..
48. Kohei Inoue, Hara Kenji, Kiichi Urahama, Integral image-based halftoning, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.67.J303, 67, 8, 2013.09, We propose a method for generating halftone images from grayscale images based on their integral images. The pixel values of a halftone image are determined so that the errors between the integral image of the halftone image and that of the grayscale image are minimized..
49. Kenji Hara, Kohei Inoue, kiichi urahama, Generalized Mixture Ratio Based Blind Image Separation, IEEE Signal Processing Letters (SPL), 10.1109/LSP.2013.2265274, vol.20, no.8, pp.743-746, 2013.08.
50. Kohei Inoue, Hara Kenji, Kiichi Urahama, Content-aware image resizing based on row/column clustering, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.67.J45, 67, 2, 2013.02, We propose an image resizing method using row/column removal and addition based on the clustering of rows and columns in an image. To shrink an image, we first separately cluster a set of rows and a set of columns in the original image and then replace each cluster with a representative row or column to reduce the numbers of rows and columns. To enlarge an image, we add linearly interpolated rows and columns between adjacent clusters. Experimental results show that the proposed method can shrink or enlarge given images while preserving the aspect ratios of the main objects in the images..
51. Kohei Inoue, Hara Kenji, Kiichi Urahama, Minimum spanning tree filters for edge-preserving impulse noise removal, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.66.J287, 66, 8, 2012.08, We propose edge-preserving smoothing (EPS) filters based on minimum spanning trees (MSTs) on image planes. The shape of the filter window is adaptively determined for each pixel by the MST, which avoids stepping over edges in images. Experimental results showed that the proposed MST filters can remove impulse noise better than conventional EPS filters..
52. Kohei Inoue, Hara Kenji, Kiichi Urahama, Contrast enhancement of color images with boundary colors in RGB color space, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.66.J227, 66, 7, 2012.07, We propose a method for enhancing the contrast of color images by transforming the original colors into the colors on the boundary of RGB color space. Experimental results show that the proposed method generates colorful and contrast-enhanced images..
53. Hara Kenji, Kohei Inoue, Kiichi Urahama, Resolving permutation ambiguity in correlation-based blind image separation, Pattern Recognition Letters, 10.1016/j.patrec.2011.11.011, 33, 5, 559-567, 2012.04, We address the problem of permutation ambiguity in blind separation of multiple mixtures of multiple images (resulting, for instance, from multiple reflections through a thick grass plate or through two overlapping glass plates) with unknown mixing coefficients. In this paper, first we devise a generalized multiple correlation measure between one gray image and a set of multiple gray images and derive a decorrelation-based blind image separation algorithm. However, many blind image separation methods, including this algorithm, suffer from a permutation ambiguity problem that the success of the separation depends upon the selection of permutations corresponding to the orders of the update operations. To solve the problem, we improve the first algorithm above by decorrelating the mixtures while searching for the appropriate update permutation using a pruning technique. We show its effectiveness through experiments with artificially mixed images and real images..
54. Kohei Inoue, Kenji Hara, Kiichi Urahama, Symmetric generalized low rank approximations of matrices, Proc. of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP2012), 2012.03.
55. Kenji Hara, Kohei Inoue, Kiichi Urahama, Resolving Permutation Ambiguity in Correlation-Based Blind Image Separation, Pattern Recognition Letters (PRL), vol. 33, pp. 559-567, 2012.01.
56. Kohei Inoue, Hara Kenji, Kiichi Urahama, Saliency map based on color histograms of inside and outside of clipped rectangular region, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.65.996, 65, 7, 996-999, 2011.12, We have proposed a method for clipping a rectangular region from an image by minimizing a weighted intersection of two color histograms that are constructed with pixels inside and outside the rectangular region. We propose a method for detecting salient regions which is based on two color histograms obtained by the clipping. We experimentally compared our method with the latest method by Achanta et al., and the results show that salient regions can be detected more clearly with our method..
57. Kohei Inoue, Hara Kenji, Kiichi Urahama, Seam carving using color saliency map, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.65.1003, 65, 7, 1003-1006, 2011.12, We describe a method for reducing an image by seam carving using a color saliency map in which higher values are assigned to the pixels with the object's colors. Our experimental results show that the proposed method can preserve the salient regions in the reduced images better than the conventional methods..
58. Kohei Inoue, Kenji Hara, Kiichi Urahama, Non-Iterative Symmetric Two-Dimensional Linear Discriminant Analysis, IEICE Transactions on Information and Systems, Vol.E94-D,No.4,pp.-, 2011.04.
59. Kenji Hara, Ko Nishino, Variational Estimation of Inhomogeneous Specular Reflectance and Illumination from A Single View, OSA Journal of Optical Society of America A (JOSA A), vol. 28, no. 2, 2011.02, [URL].
60. Kohei Inoue, Hara Kenji, Kiichi Urahama, Non-iterative symmetric two-dimensional linear discriminant analysis, IEICE Transactions on Information and Systems, 10.1587/transinf.E94.D.926, E94-D, 4, 926-929, 2011.01, Linear discriminant analysis (LDA) is one of the wellknown schemes for feature extraction and dimensionality reduction of labeled data. Recently, two-dimensional LDA (2DLDA) for matrices such as images has been reformulated into symmetric 2DLDA (S2DLDA), which is solved by an iterative algorithm. In this paper, we propose a non-iterative S2DLDA and experimentally show that the proposed method achieves comparable classification accuracy with the conventional S2DLDA, while the proposed method is computationally more efficient than the conventional S2DLDA..
61. Hara Kenji, Ko Nishino, Variational estimation of inhomogeneous specular reflectance and illumination from a single view, Journal of the Optical Society of America A: Optics and Image Science, and Vision, 10.1364/JOSAA.28.000136, 28, 2, 136-146, 2011.01, Estimating the illumination and the reflectance properties of an object surface from a few images is an important but challenging problem. The problem becomes even more challenging if we wish to deal with real-world objects that naturally have spatially inhomogeneous reflectance. In this paper, we derive a novel method for estimating the spatially varying specular reflectance properties of a surface of known geometry as well as the illumination distribution of a scene from a specular-only image, for instance, recovered from two images captured with a polarizer to separate reflection components. Unlike previous work, we do not assume the illumination to be a single point light source. We model specular reflection with a spherical statistical distribution and encode its spatial variation with a radial basis function (RBF) network of their parameter values, which allows us to formulate the simultaneous estimation of spatially varying specular reflectance and illumination as a constrained optimization based on the I-divergence measure. To solve it, we derive a variational algorithm based on the expectation maximization principle. At the same time, we estimate optimal encoding of the specular reflectance properties by learning the number, centers, and widths of the RBF hidden units. We demonstrate the effectiveness of the method on images of synthetic and real-world objects..
62. Kohei Inoue, Kenji Hara, Kiichi Urahama, Non-Iterative Symmetric 2DLDA, 2nd China-Japan-Korea Joint Workshop on Pattern Recognition (CJKPR), 2010.11.
63. Kohei Inoue, Kenji Hara, Kiichi Urahama, Image and Video Clipping by Weighted Histogram Intersection Minimization, TENCON, 2010.11.
64. Kohei Inoue, Kenji Hara, Kiichi Urahama, Image Resizing by Row/Column Removal and Addition Based on Hierarchical Clustering, TENCON, 2010.11.
65. Kenji Hara, Kohei Inoue, Kiichi Urahama, Pruning-Based Blind Separation of Image Mixtures, Proc. of IEEE International Conference on Image Processing (ICIP2010), accepted, 2010.10.
66. Real-Time Nonlinear FEM-Based Simulator for Deforming Volume Model of Soft Organ by Neural Network.
67. Kohei Inoue, Kenji Hara, Kiichi Urahama, Image Clipping by Weighted Histogram Intersection Minimization, 16th Korea-Japan Joint Workshop on Frontiers of Computer Vision (FCV2010), to appear, 2010.02.
68. Daisuke Miyazaki, Hara Kenji, Katsushi Ikeuchi, Median photometric stereo as applied to the segonko tumulus and museum objects, International Journal of Computer Vision, 10.1007/s11263-009-0262-9, 86, 2-3, 229-242, 2010.01, One of the necessary techniques for constructing a virtual museum is to estimate the surface normal and the albedo of the artwork which has high specularity. In this paper, we propose a novel photometric stereo method which is robust to the specular reflection of the object surface. Our method can also digitize the artwork arranged inside a glass or acrylic display case without bringing the artwork out of the display case. Our method treats the specular reflection at the object surface or at the display case as an outlier, and finds a good surface normal evading the influence of the outliers. We judiciously design the cost function so that the outlier will be automatically removed under the assumption that the object's shape and color are smooth. At the end of this paper, we also show some archived 3D data of Segonko Tumulus and objects in the University Museum at The University of Tokyo that were generated by applying the proposed method..
69. Trachea and Esophagus Classification by AdaBoost.
70. Kohei Inoue, Hara Kenji, Kiichi Urahama, Entropy enhancement for effective width reduction in panoramic images, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.63.1622, 63, 11, 1622-1624, 2009.12, We have proposed a method for reducing the width of panoramic images with high aspect ratios on the basis of the entropy in local color histograms computed for each column in the images. In this paper, we propose a method for enhancing the entropy in the conventional method to further widen and narrow the regions of high and low entropies, respectively. Experimental results show that the proposed method can reduce the width of panoramic images more effectively than the conventional method..
71. Kohei Inoue, Kenji Hara, Kiichi Urahama, Robust Multilinear Principal Component Analysis, Proc. of IEEE International Conference on Computer Vision (ICCV2009), to appear, 2009.10.
72. Kenji Hara, Ko Nishino, Illumination and Spatially Varying Specular Reflectance from a Single View, Proc. of IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR2009), to appear, 2009.06.
73. Daisuke Miyazaki, Kenji Hara, Katsushi Ikeuchi, Median Photometric Stereo as Applied to the Segonko Tumulus and Museum Objects, International Journal of Computer Vision (IJCV), to appear, 2009.06.
74. Kenji Hara, Kohei Inoue, Kiichi Urahama, Separation of Layers from Images containing Multiple Reflections and Transparency using Cyclic Permutation, Proc. of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP2009), to appear, 2009.04.
75. Chang Jian, Kohei Inoue, Hara Kenji, Kiichi Urahama, Fixed-coefficient iterative bilateral filters for graph-based image processing, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 10.1007/978-3-540-92957-4_41, 5414 LNCS, 473-484, 2009.02, We present a graph-based image processing algorithm using fast iterative bilateral filters. The computation of bilateral filters is accelerated with fixation of the coefficients during iterations and their approximate decomposition further speeds up the computation. We show that this fixed-coefficient iterative bilateral filter is an alternative solver for optimization problems in graph-based data analyses and apply its fast algorithm to graph-based image processing tasks. Performance of the present algorithm is demonstrated with experiments of contrast enhancement and smoothing of images using cross bilateral filters, in addition to semi-supervised image segmentation and colorization of monochromatic images..
76. Chang Jian, Kohei Inoue, Kenji Hara, Kiichi Urahama, Fixed-Coefficient Iterative Bilateral Filters for Graph-Based Image Processing, Proc. of IEEE Pacific-Rim Symposium on Image and Video Technology (PSIVT2009), to appear, 2009.01.
77. Kohei Inoue, Kenji Hara, Kiichi Urahama, Robust Simultaneous Low Rank Approximation of Tensors, Proc. of IEEE Pacific-Rim Symposium on Image and Video Technology (PSIVT2009), to appear, 2009.01.
78. Ken'ichi Morooka, Xian Chen, Ryo Kurazume, Seiichi Uchida, Kenji Hara, Yumi Iwashita, Makoto Hashizume, Real-time Nonlinear FEM with Neural Network for Simulating Soft Organ Model Deformation, Proc. of the 11th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2008),, to appear, 2008.09.
79. 2D/3D Registration by Back Projection and Geometrical Constraints.
80. Kenji Hara, Ko Nishino, Katsushi Ikeuchi, Mixture of Spherical Distributions for Single-View Relighting, IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), vol.30, no.1, pp.25-35, 2008.01.
81. Hara Kenji, Ko Nishino, Katsushi Ikeuchi, Mixture of spherical distributions for single-view relighting, IEEE Transactions on Pattern Analysis and Machine Intelligence, 10.1109/TPAMI.2007.1164, 30, 1, 25-35, 2008.01, We present a method for simultaneously estimating the illumination of a scene and the reflectance property of an object from single view images - a single image or a small number of images taken from the same viewpoint. We assume that the illumination consists of multiple point light sources and the shape of the object is known. First, we represent the illumination on the surface of a unit sphere as a finite mixture of von Mises-Fisher distributions based on a novel spherical specular reflection model that well approximates the Torrance-Sparrow reflection model. Next, we estimate the parameters of this mixture model including the number of its component distributions and the standard deviation of them, which correspond to the number of light sources and the surface roughness, respectively. Finally, using these results as the initial estimates, we iteratively refine the estimates based on the original Torrance-Sparrow reflection model. The final estimates can be used to relight single-view images such as altering the intensities and directions of the individual light sources. The proposed method provides a unified framework based on directional statistics for simultaneously estimating the intensities and directions of an unknown number of light sources as well as the specular reflection parameter of the object in the scene..
82. Kenichi Morooka, Xian Chen, Ryo Kurazume, Seiichi Uchida, Hara Kenji, Yumi Iwashita, Makoto Hashizume, Real-time nonlinear FEM with neural network for simulating soft organ model deformation., Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention, 11, Pt 2, 742-749, 2008.01, This paper presents a new method for simulating the deformation of organ models by using a neural network. The proposed method is based on the idea proposed by Chen et al. that a deformed model can be estimated from the superposition of basic deformation modes. The neural network finds a relationship between external forces and the models deformed by the forces. The experimental results show that the trained network can achieve a real-time simulation while keeping the acceptable accuracy compared with the nonlinear FEM computation..
83. Kenji Hara, Ryo Kurazume, Kohei Inoue, Kiichi Urahama, Segmentation of Images on Polar Coordinate Meshes, Proc. of IEEE International Conference on Image Processing(ICIP2007), pp.245-248, 2007.09.
84. Fast 3D Shape Reconstruction of Moving Object by Parallel Fast Level Set Method.
85. Daisuke Makino, Kohei Inoue, Hara Kenji, Kiichi Urahama, Generating paper mosaic images taking into account thickness and texture of paper, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.60.2038, 60, 12, 2038-2041, 2006.12, We describe a non-photorealistic rendering method for generating paper mosaic images that takes into account and compensates for the effects of the thickness and texture of the paper onto which the image is projected. We modulate the brightness of the paper mosaic image on the basis of the emboss-filtered image of the overlapped pieces of paper. We also modulate the brightness of the paper mosaic image on the basis of the paper's texture. That is, we incorporate the effects of the paper's thickness and texture into the resultant paper mosaic image. We experimentally showed that our method increases the three-dimensional sense of depth in paper mosaic images..
86. Shuntaro Yui, Hara Kenji, Hongbin Zha, Tsutomu Hasegawa, Acceleration of narrow band method and its application to topology-adaptive 3D geometrical modeling, Systems and Computers in Japan, 10.1002/scj.20130, 37, 11, 68-78, 2006.10, A method for increasing the speed of surface shape evolution called the fast narrow band method and a topology-adaptive geometrical modeling method that uses the fast narrow band method are proposed. Although there have been many reports of the use of the level set method as an effective method for topology-adaptive geometrical modeling, the methods suffers from the problem that the computational cost is very large. The narrow band method has been widely used as the most general method for increasing the speed of the level set method, but it has not been found to sufficiently reduce the computational cost. The proposed fast narrow band method retains the precision of the narrow band method while reducing the computational cost by using distance information to increase the efficiency of the search process. Experiments are performed on scenes containing objects with holes and scenes containing multiple objects, and a five-fold increase in speed is achieved over the Hermes algorithm, which was previously the fastest implementation of the level set method..
87. Kohei Inoue, Hara Kenji, Kiichi Urahama, A two-scan algorithm for weighted Voronoi tessellations, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.60.1072, 60, 7, 1072-1076, 2006.07, We developed a two-scan algorithm for discrete Voronoi tessellations of digital images. The computation time of our method is independent of the number of Voronoi sites. In addition to the previous additively, multiplicatively, compoundly weighted Voronoi diagrams and the additively weighted power Voronoi diagram, we proposed a new weighted Voronoi diagram, namely the compoundly weighted power Voronoi diagram. These five weighted Voronoi diagrams were efficiently computed with our two-scan algorithm..
88. Kenji Hara, Kohei Inoue, Kiichi Urahama, A Factorization Method for Fish-eye Cameras, International Workshop on Advanced Image Technology, 10.3169/itej.60.1072, 759-764, 2006.
89. Kenji Hara, Ko Nishino, Katsushi Ikeuchi, Multiple Light Sources and Reflectance Property Estimation Based on a Mixture of Spherical Distributions, Proc. of IEEE International Conference on Computer Vision (ICCV2005), 1627-1634, pp.1627-1634, 2005.10.
90. Kenji Hara, Ko Nishino, Katsushi Ikeuchi, Light Source Position and Reflectance Estimation from a Single View without The Distant Illumination Assumption, IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 10.1109/TPAMI.2005.82, 27, 4, 493-505, vol.27, no.4, pp.493-505, 2005.04.
91. 3D Tracking of Multiple Moving Objects using Fast Level Set Method.
92. Daisuke Miyazaki, Robby T. Tan, Kenji Hara, Katsushi Ikeuchi, Polarization-based Inverse Rendering from a Single View: Estimating Shape, Texture, Roughness, and Illumination, Proc. of IEEE International Conference on Computer Vision (ICCV2003), pp.982-987, 2003.10.
93. Kenji Hara, Ko Nishino, Katsushi Ikeuchi, Determining Reflectance and Light Position from a Single Image without Distant Illumination Assumption, Proc. of IEEE International Conference on Computer Vision (ICCV2003), vol.I, pp.560-567, 2003.10.
94. Fast Level Set Method and Realtime Tracking of Moving Objects in a Sequence of Images.
95. K. Hara, K. Nishino, K. Ikeuchi,
"Determining Reflectance Parameters from a Single Image under Perspective Projection",
IPSJ Trans. on Computer Vision and Image Media, vol.43, no.SIG 11 (CVIM 5), pp121-129, 2002..
96. Shuntaro Yui, Kenji Hara, Hongbin Zha, Tsutomu Hasegawa, A Fast Narrow Band Method and Its Application in Topology-Adaptive 3-D Modeling, Proc. of IAPR International Conference on Pattern Recognition (ICPR'02), 122-125, vol.IV, pp.122-125, 2002.08.
97. Hara Kenji, Hongbin Zha, Tsutomu Hasegawa, Topology-adaptive modeling of objects using surface evolutions based on 3D mathematical morphology, Systems and Computers in Japan, 10.1002/scj.1152, 33, 9, 58-67, 2002.08, Level set methods were proposed mainly by mathematicians for constructing a model of a 3D object of arbitrary topology. However, those methods are computationally inefficient due to repeated distance transformations and increased dimensions. In the paper, we propose a new method of modeling fast objects of arbitrary topology by using a surface evolution approach based on mathematical morphology. Given sensor data covering the whole object surface, the method begins with an initial approximation of the object by evolving a closed surface into a model topologically equivalent to the real object. The refined approximation is then performed using energy minimization. The method has been applied in several experiments using range data, and the results are reported in the paper..
98. Shuntaro Yui, Hara Kenji, Hongbin Zha, Tsutomu Hasegawa, A fast narrow band method and its application in topology-adaptive 3-D modeling, 16th IAPR International Conference on Pattern Recognition, 16, 4, 122-125, 2002, We present a new fast method of modeling 3-D objects of arbitrary topology. The Level Set Methods have been used by many researchers to recover 3-D shapes of arbitrary topology. However, those methods are computationally inefficient. To reduce the computational cost, a new method named FNB (Fast Narrow Band Method) is proposed. FNB is based on the Narrow Band Method (NB) which is the well-known fast method of the Level Set Method. The main idea is to exploit the combinative use of the narrow band and the approximate distance from the front. The method has been applied in several experiments using range data and we find that our method is over 50 times faster than NB..
99. Hara Kenji, Ko Nishino, Katsushi Ikeuchi, Estimation of Illuminant Position and Reflectance Properties Using Polarization under Perspective Projection, IAPR Workshop on Machine Vision Applications, 1, 566-571, 2002.
100. Hara Kenji, H. Zha, T. Hasegawa, Surface Evolution Based on Wave Propagation and Its Application to Topology-Adaptive 3-D Modeling, IAPR Workshop on Machine Vision Applications, 1, 87-90, 2000.
101. N. Fujii, M. Fujitake, Hara Kenji, Two-dimensional motion motor with circular core and plural divided windings supplied separately, IEEE Transactions on Magnetics, 10.1109/20.800738, 35, 5 PART 2, 4010-4012, 1999.12, Abstract - We have proposed a circular shape of induction type two-dimensional motion motor called the surface induction motor in which the winding is controlled at each coil. For a simpler drive using conventional three-phase inverters, a surface Induction motor with divided windings is proposed, in which one inverter is used for supplying power to one group of windings. Continuously change of thrust direction is verified by using a test machine. The experimental and theoretical results give that the six-division type with three three-phase inverters or eight-division type surface motor supplied by four threephase inverters is best..
102. N. Fujii, M. Fujitake, Hara Kenji, Two-dimensional motion motor of circular core with plural divided windings supplied separately, Quaternary International, 1999, The surface induction motor with plural divided windings is proposed to simplify the basic drive method. The characteristics of this type motor are demonstrated both experimentally and theoretically..
103. K. Hara, H. Zha, T. Hasegawa, 3-D Free-Form Object Description and Modeling Using Real Range Image, 4th Japan-France / 2nd Asia-Europe Congress on Mechatronics, 3, 578-583, 1998.
104. Hara Kenji, H. Zha, T. Hasegawa, T. Nagata, 3-D Object modeling based on surface reconstruction and integration using optimal polygonal approximation, Unknown Journal, 3, 2644-2649, 1997, In the paper, we address the problem of 3-D object modeling that is suitable for graphic display by deriving an optimal approximation to the true object surfaces. A new model generation method is proposed, which obtains a mesh representation of the surfaces by integrating multiple range images with an optimization algorithm based on closeness to the original image data. The method is a relaxation algorithm that achieves accurate shape descriptions gradually by modifying and propagating mesh arrangements iteratively. The paper first presents the formulation within the framework of regularization by redefining the minimization energy function in which both the data closeness and surface smoothness are considered. Next, we give the detailed procedure of the method repeating the surface reconstruction using the chosen regularization parameter at each stage. The search vector field corresponding to range data is generated alternately. The method has been applied to real range images to verify the applicability, and results are reported in the paper..
105. Hara Kenji, H. B. Zha, T. Nagata, 3-D Shape Reconstruction by Using Statistical Models, 1996 IEEE International Conference on Systems, Man, and Cybernetics, 2, 1097-1102, 1996.
106. Reconstruction of Contour Features and 3-D Surfaces by Using a Statistical Model.
107. Hara Kenji, H. B. Zha, T. Nagata, Curve and surface reconstruction by using sequential Markov random fields (MRFs), 1995 IEEE International Conference on Systems, Man, and Cybernetics, 2, 1097-1102, 1995, The statistical approach using the coupled Markov random field (MRF) and the maximum a posteriori (MAP) estimate has been proposed in order to satisfy both the preservation of local discontinuities and the smoothing of continuous regions for reconstruction of derivative feature measurements and range data. However, this data reconstruction method has some very difficult problems that it is hard to obtain proper results by finding the global optimal solution correctly. Especially, if the ordinary iteration solutions are used for the noisy and rugged data, is it likely that the convergence happens to be at a local optimal solution depending upon the initial value, and the noise smoothing is insufficient or the edge parts are overslurred. To cope with the difficulties, we propose a recovery method that regards the MAP estimation itself as the basic process and do the computation iteratively while controlling smoothing by changing values of the MRF parameters according to some scheduling. The presented algorithm reduces failures because of the local optimization, and is respected to give better results of reconstruction. The applicability of the method has been verified by several reconstruction experiments..
108. Hara Kenji, H. B. Zha, T. Nagata, Contour Segmentation by Using an Extended Koch Neural Network, International Conference on Automation, Robotics and Computer Vision, 2, 1309-1313, 1994.