KINDAI UNIVERSITY


*A space between the first name and last name, please enter

YOSHIDA Hiromi

Profile

FacultyDepartment of Electronic Engineering and Computer Science
PositionLecturer
DegreePh.D. in Engineering
Commentator Guidehttps://www.kindai.ac.jp/meikan/2334-yoshida-hiromi.html
URL
Mail
Last Updated :2020/09/02

Research Activities

Research Areas

  • Informatics, Software

Published Papers

  • Low-Light Image Enhancement via Adaptive Shape and Texture Prior, Kazuki Kurihara, Hiromi Yoshida, Youji Iiguni, The 15th International Conference on SIGNAL IMAGE TECHNOLOGY & INTERNET BASED SYSTEMS, The 15th International Conference on SIGNAL IMAGE TECHNOLOGY & INTERNET BASED SYSTEMS, 74 - 81, Nov. 2019 , Refereed
  • Dehazing with Recovery Level Map: Suppressing Over-Enhancement and Residual Haze, Kentaro Iwamoto, Hiromi Yoshida, Youji Iiguni, 67 - 73, Nov. 2019 , Refereed
  • Pencil Art Generation Method Focusing on Pencil Line Expression, 吉田 大海, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 47(1), 41 - 51, 2018 , Refereed
  • Restoration of Wiper Region Using Mask Processing for White Line Detection, 中嶋 研生, 吉田 大海, 飯國 洋二, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 47(2), 167 - 174, 2018 , Refereed
  • A Pencil Texture Generation Based on LIC Method and Analog Technique, Yoshida Hiromi, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 46(3), 395 - 402, 2017 , Refereed
    Summary:This paper proposes a method for generating digital pencil drawing texture from grayscale image. The proposed method introduces analog techniques of pencil drawing expressly by using hatching textures which can be generated based on LIC method. Analog techniques introcuded by proposed method are hatching, cross-hatching, and blending. Hatching is a technique of generating texture by drawing multiple parallel line segments. Cross-hatching can be generated by stacking hatchings in layers, which gives the effect of sensitive contrasting density to pencil drawing. Blending can emphasize the effects of cross-hatching by rubbing it with cloth or finger. Some of conventional methods of generating pencil drawing also introduce analog techniques like these, which do not use them expressly. This paper also confirms the effect of introducing analog techniques of pencil drawing expressly. LIC method can generate hatching which is a basic element of these analog techniques expressly and effectively. So, proposed method generates hatching textures differ from each other in luminance and direction based on LIC method first. And pencil drawing textures which expressly reflects analog techniques of pencil drawing are realized based on total product.
  • A Pencil Texture Generation Based on LIC Method and Analog Technique, 吉田 大海, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 画像電子学会誌 The journal of the Institute of Image Electronics Engineers of Japan : visual computing, devices & communications, 46(3), 395 - 402, 2017 , Refereed
  • Automated GrowCut with Multilevel Seed Strength Value for Building Image, 上野 大, 吉田 大海, 飯國 洋二, システム制御情報学会論文誌 = Transactions of the Institute of Systems, Control and Information Engineers, システム制御情報学会論文誌 = Transactions of the Institute of Systems, Control and Information Engineers, 29(6), 266 - 274, Jun. 2016 , Refereed
  • Pencil Line Drawing Based on Size of Graphic, Yoshida Hiromi, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 45(2), 212 - 215, 2016 , Refereed
    Summary:This paper describes an improvement of pencil stroke map which is proposed as a pencil line drawing method. Pencil stroke map detects gradient of image and products pencil line drawing by convolution line segment with respect to each angle of gradient. However, this method has the problem that if length of line segment for convolution is unsuitable for size of graphic, resulting image loses the shape of them or degrades its quality as pencil line drawing. On this account, proposed method classifies graphics by size and change the length of line segment depending on classification result, resulting image can be obtained with improved this problem. In experiment, this paper shows effectiveness of proposed method by comparison of resulting image between proposed method and pencil stroke map.
  • Virtual Wiper-Restoration from Vehicle-Mounted Camera Using Wiper Template, Akio Nakashima, Hiromi Yoshida, Youji Iiguni, Proceedings of the International Symposium on ENGINEERING SCIENCE 2015, Proceedings of the International Symposium on ENGINEERING SCIENCE 2015, 256 - 260, May 2015 , Refereed
  • Advanced Gabor Filter for Low-quality Document Image Based on Visual Feature, Satoshi Yoneda, Hiromi Yoshida, Youji Iiguni, Proceedings of the World Engineering Conference and Convention 2015, Proceedings of the World Engineering Conference and Convention 2015, 1 - 5, 2015 , Refereed
  • Image Inpainting Method Based on Painting Texture, Yoshida Hiromi, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 44(1), 77 - 84, 2015 , Refereed
    Summary:In this paper, I propose an image inpainting technique for damaged oil painting that aims to create their restored examples. Usually, image inpainting technique has the advantage of being able to work well for a wide variety of images because it does not require any information about target images. However, image inpainting has the potential to work more effectively if it limits the targets and introduces the processing based on property of oil painting. Proposed method achieves high performance for oil painting images by introducing two processing based on property of them. One is to apply smoothing filter based on painting technique, and the other is to add asperity of painting as texture.
  • Scratch Inpainting Method Based on Space Distance, Yoshida Hiromi, Iiguni Youji, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 43(3), 374 - 382, 2014 , Refereed
    Summary:This paper proposes inpainting technique based on space distance for scratched image which damage location is given. Conventional methods using pixel value can be categorized into two major classes. One is a pixel-based method which interpolates missing regions on a pixel-to-pixel basis by evaluating continuity of luminance. The other is a patch-based method which uses image fragment for interpolation based consistency between the missing region and source region. However, if scratch damages are widely and densely distributed on the image, both methods are inapplicable to it. In the context of it, the inpainting technique which considers space distance between the missing region and source region as a likelihood is proposed. By the reason that the proposed method is categorized into a pixel-based method, it is applicable to images which can not secure the image fragment required by patch-based method. In addition, interpolating without having to refer the pixel values of source region, it is also applicable to images which can not evaluate continuity of luminance. Since the interpolation pixels are determined by only it's location in missing region, the proposed method requires no iterative computation and the information of mid-flow, resulting in high-speed computing. The experiment shows performance of proposed method for variety of images.
  • A Draught Reading Method by Image Processing with the Robustness of Measurement Distance, OKAMOTO Akihiro, YOSHIDA Hiromi, TANAKA Naoki, HIRONO Kouhei, The Journal of Japan Institute of Navigation, The Journal of Japan Institute of Navigation, 130(0), 135 - 140, 2014 , Refereed
    Summary:It is an important step to read draughts accurately in the process of draught surveys because inaccurate one could lead to a cause of disputes between all the interested parties. The draught reading is conducted by ship's officers or professional surveyors based on their visual observation even today. Therefore, there are some difficulties in keeping an accurate measurement or observer's safety under unfavorable conditions such as bad weather, strong swell and so forth. In this paper, the authors propose a novel draught reading method using image processing for supporting reading draughts. The method comprises some processing steps, detecting draught marks, stabilizing the hull motion and recognizing draught marks by template matching, and detecting waterline due to accumulated sequential inter-frame differences. Experimental results show that the proposed method attained high precision of the measurement.
  • A Laser Image Binarization for Crack Detection by Fractal Dimension, Yoshida Hiromi, Iiguni Youji, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 43(4), 595 - 598, 2014 , Refereed
    Summary:This paper proposes laser image binarization for crack detection by fractal dimension. Laser image is one of the road surface image and can be obtained by observing scattering light of laser from road surface. It has the advantage of robustness of lighting condition over an image that is taken by optical camera. However, surface asperity of road manifests as texuture regions in these images. So, crack detection method needs to be deal with texture regions. Although many of conventional methods geared toward concrete construct, these methods do not work well to road surface image. Proposed method can evaluate a texture region and binarized laser image effectively through the use of blanket method locally and globally.
  • A Conditional Image Enhancement for the Low-quality Document Image, Yoshida Hiromi, The Journal of the Institute of Image Electronics Engineers of Japan, The Journal of the Institute of Image Electronics Engineers of Japan, 43(1), 28 - 34, 2014 , Refereed
    Summary:A conditional image enhancement for the low-quality document image is proposed. In a document recognition by software such as OCR, image resolution is very important factor for recognition rate and an images which are low-resolution can be expected to be improved through the existing enhancement methods which can add high-frequency component. However, if images include such noise regions as bleed-through, blot and slur, they process not only improving character regions but also increasing a noises, and cause false recognition. The proposed method can improve the character regions without noise enhancement by conditional adding high-frequency component. This paper shows the effective using OCR software and by comparing existing enhancement method.
  • A New Method of Character Strings Extraction based on Blanket Binarization, Hiromi Yoshida, Naoki Tanaka, Takashi Nagamatsu, International Workshop on Document Analysis Systems, Short Paper Booklet, International Workshop on Document Analysis Systems, Short Paper Booklet, 9 - 10, May 2013 , Refereed
  • A Binarization Method for Degraded Document Images with Morphological Operations., Akihiro Okamoto, Hiromi Yoshida, Naoki Tanaka, Proceedings of the 13. IAPR International Conference on Machine Vision Applications, MVA 2013, Kyoto, Japan, May 20-23, 2013, Proceedings of the 13. IAPR International Conference on Machine Vision Applications, MVA 2013, Kyoto, Japan, May 20-23, 2013, 294 - 297, 2013 , Refereed
  • An Improvement of Compression Algorithm of Weather Chart Suited for Transmission by Voice Band of Fisheries Radio, Yoshida Hiromi, Uemura Hiroyuki, Tanaka Naoki, Inoue Takeshi, The Journal of Japan Institute of Navigation, The Journal of Japan Institute of Navigation, 124(0), 345 - 353, 2011 , Refereed
    Summary:Transmitting a weather chart on a fisheries radio voice band, data size of a chart must be less than 3KB in order to keep the required time of a few minutes. Since it has over 340KB data size originally, we need to reduce the size by one hundredth. In this paper, we propose a new compression method for weather chart. Compared with the previous method, proposed method includes the following advantages. Implementation of error detection and feedback system, reducing computational cost by speeding up blanket method, and improvement of visibility by filtering process based on character features.
  • A Blanket Binarization Method for Character String Extraction., Hiromi Yoshida, Naoki Tanaka, Proceedings of the IAPR Conference on Machine Vision Applications (IAPR MVA 2011), Nara Centennial Hall, Nara, Japan, June 13-15, 2011, Proceedings of the IAPR Conference on Machine Vision Applications (IAPR MVA 2011), Nara Centennial Hall, Nara, Japan, June 13-15, 2011, 31 - 34, 2011 , Refereed
  • A Binarization Method for the Character Extraction from Scene Image by the Blanket Method, Hiromi Yoshida, Naoki Tanaka, Journal of the Institute of Image Electronics Engineers of Japan, Journal of the Institute of Image Electronics Engineers of Japan, 39(4), 463 - 472, 2010 , Refereed
    Summary:We propose a new global binarization method suited for character strings extraction from a scene image. To obtain a high-quality segmentation result, binarization is a significant step in character extraction process. Since character strings region of sign board has the wide variety of layout and color, it is hard to binarize this category of images by conventional global methods, which are based on statistical information. The proposed method can select a threshold by evaluating the fineness of the binarized image by using the fractal dimension. By evaluating the compactness and stability of the regions of binarized image with the fractal dimension calculated by blanket method, the proposed method achieves fine binarzation results, which are no way inferior to manually and/or using one of promising global methods known as Otsu binarizing method. © 2010, The Institute of Image Electronics Engineers of Japan. All rights reserved.
  • A New Compression Algorithm of Weather Chart Suited for Transmission by Voice Band of Fisheries Radio, Yoshida Hiromi, Tanaka Naoki, Inoue Takeshi, The Journal of Japan Institute of Navigation, The Journal of Japan Institute of Navigation, 122(0), 61 - 67, 2010 , Refereed
    Summary:Transmitting a weather chart on a fisheries radio voice band, data size of a chart must be less than 3KB in order to keep the required time of a few minutes. Since it has over 340KB data size originally, we need to reduce the size by one hundredth. With conventional method, e.g. FAX format, 11KB is the best compressed size. In this paper we propose a new compression method for weather chart which achieves 3KB of compressed size with a fine quality of output image.
  • On Misuse of the Connectivity Condition in the Hilditch Thinning Algorithm, Hiromi Yoshida, Naoki Tanaka, Takeshi Inoue, Journal of the Institute of Image Electronics Engineers of Japan, Journal of the Institute of Image Electronics Engineers of Japan, 39(4), 490 - 493, 2010 , Refereed
    Summary:The Hilditch thinning algorithm is a sophisticated method for digital images and thought to be still useful in today. A kind of faulty version of Hilditch thinning algorithm which acts on incorrect connectivity condition has been found in some theses and/or on web sites. In this paper, we point out why and how the false version has appeared to request attention to algorithm details to use a strict Hilditch thinning algorithm. © 2010, The Institute of Image Electronics Engineers of Japan. All rights reserved.
  • A Binarization Method for Crack Detection in a Road Surface Image with the Fractal Dimension., Hiromi Yoshida, Naoki Tanaka, Proceedings of the IAPR Conference on Machine Vision Applications (IAPR MVA 2009), Keio University, Yokohama, Japan, May 20-22, 2009, Proceedings of the IAPR Conference on Machine Vision Applications (IAPR MVA 2009), Keio University, Yokohama, Japan, May 20-22, 2009, 70 - 73, 2009 , Refereed
  • A Study on Signboard Image Identification with SIFT Features, In Handout of 8th IAPR International Workshop on Document Analysis System, In Handout of 8th IAPR International Workshop on Document Analysis System, 10 - 13, 2008 , Refereed
  • Automatic pencil sketch generation by using canny edges, Ryota Okawa, Hiromi Yoshida, Youji Iiguni, Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017, Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017, 282 - 285, Jul. 19 2017 , Refereed
    Summary:This paper presents a system that automatically converts 2D raster images to sketch style. The proposed method first extracts edges at different resolutions. Then, these shapes and brightness are varied and merged. This process expresses trial and error in the actual sketch. Experimental results showed that the proposed method produces images with natural appearance.
  • Detection of view-disturbing noise by using time-axial clustering in spatio-temporal image, Taichi Arimasa, Hiromi Yoshida, Yoji Iiguni, Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017, Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017, 278 - 281, Jul. 19 2017 , Refereed
    Summary:Lately, surveillance cameras are widely utilized. Nevertheless, raindrops or mud sometimes obstruct the field of view when we use them outside. In this research, we propose a method that can remove noises better than conventional method. In the conventional method, we take photographs while rotating the camera. Then, by considering the relationship between angles of photographing and objects in the images, we can detect noises. But in this method, if the luminance of the objects and noises are so close, we cannot detect them exactly. So we solve the problem by using clustering.
  • Automatic draft reading based on image processing, Takahiro Tsujii, Hiromi Yoshida, Youji Iiguni, OPTICAL ENGINEERING, OPTICAL ENGINEERING, 55(10), 104104 - 104104, Oct. 2016 , Refereed
    Summary:In marine transportation, a draft survey is a means to determine the quantity of bulk cargo. Automatic draft reading based on computer image processing has been proposed. However, the conventional draft mark segmentation may fail when the video sequence has many other regions than draft marks and a hull, and the estimated waterline is inherently higher than the true one. To solve these problems, we propose an automatic draft reading method that uses morphological operations to detect draft marks and estimate the waterline for every frame with Canny edge detection and a robust estimation. Moreover, we emulate surveyors' draft reading process for getting the understanding of a shipper and a receiver. In an experiment in a towing tank, the draft reading error of the proposed method was < 1 cm, showing the advantage of the proposed method. It is also shown that accurate draft reading has been achieved in a real-world scene. (C) The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License.
  • A New Binarization Method for a Sign Board Image with the Blanket Method, Hiromi Yoshida, Naoki Tanaka, PROCEEDINGS OF THE 2009 CHINESE CONFERENCE ON PATTERN RECOGNITION AND THE FIRST CJK JOINT WORKSHOP ON PATTERN RECOGNITION, VOLS 1 AND 2, PROCEEDINGS OF THE 2009 CHINESE CONFERENCE ON PATTERN RECOGNITION AND THE FIRST CJK JOINT WORKSHOP ON PATTERN RECOGNITION, VOLS 1 AND 2, 845 - 848, 2009 , Refereed
    Summary:We propose a new binarization method suited for character extraction from a sign board in a scenery image. The binarization is one of a significant step in character extraction in order to get high quality result. Character region on sigh board, however, has many variation and colors. In addition to it, if there exists high frequency texture region like a mountain or trees in the background, it can be a cause of difficulty to binarize an image. At the high frequency region, the binarized result is sensitive to the threshold value. On the other hand, a character region of sign board consists of solid area, that is, includes few high frequency regions, and has relatively high contrast. So the binarized result of character region is stable at an interval of the threshold value. Focusing attention on this point, we propose a new binarization method which obtains a threshold value based on the fractal dimension by evaluating both region's density and stability to threshold value. Through the proposed method, we can get a fine quality binarized images, where the characters can be extracted correctly.
  • A Binarization Method for a Scenery Image with the Fractal Dimension, Hiromi Yoshida, Naoki Tanaka, PROGRESS IN PATTERN RECOGNITION, IMAGE ANALYSIS, COMPUTER VISION, AND APPLICATIONS, PROCEEDINGS, PROGRESS IN PATTERN RECOGNITION, IMAGE ANALYSIS, COMPUTER VISION, AND APPLICATIONS, PROCEEDINGS, 5856, 29 - 36, 2009 , Refereed
    Summary:We propose a new binarization method suited for character extraction from a sign board in a scenery image. The binarization is thought to be a significant step in character extraction in order to get high quality result. Character region of sigh board, however, has many variation and colors. In addition to it, if there exists high frequency texture region like a mountain or trees in the background, it can be a cause of difficulty to binarize an image. At the high frequency region, the binarized result is sensitive to the threshold change. On the other hand, a character region of sign board consists of solid area, that is, includes few high frequency regions, and has relatively high contrast. So the binarized result of character region is stabile against an interval of the threshold value. Focusing attention on this point, we propose a new method which obtains a threshold value based on the fractal dimension to evaluate both region's density and stability to threshold change. Through the proposed method, we can get a fine quality binarized images, where the characters can be extracted correctly.

Conference Activities & Talks

  • Tetrahedron Height Minimization for Vanishing Point Detection,   2019 10
  • Underwater Image Restoration Based on Luminance Map,   2019 10
  • Shadow Removal in Images Using Multi-Scale Image Decomposition and Region Segmentation,   2019 10
  • Image Enhancement By using high dynamic range and Laplacian pyramid, 川邊 彪, 吉田 大海, 飯國 洋二, 電子情報通信学会技術研究報告 = IEICE technical report : 信学技報,   2017 10 05
  • Evolutionary Creation of Painting Images For Non-Photorealistic and Textured Elements, 加藤 洋明, 吉田 大海, 飯國 洋二, 電子情報通信学会技術研究報告 = IEICE technical report : 信学技報,   2017 10 05
  • Tsunami Inundation Forecast Hazard Map Creation Method Which Is Easily Updated Using Land Cover Analysis, 中西 創, 吉田 大海, 飯國 洋二, 電子情報通信学会技術研究報告 = IEICE technical report : 信学技報,   2017 10 05
  • Automatic Generation of Pencil-like Rough Sketchy Image in Consideration of Omitting Details and Repeating Fluctuated Strokes, 松本 光司, 吉田 大海, 飯國 洋二, 電子情報通信学会技術研究報告 = IEICE technical report : 信学技報,   2016 07 01
  • Seamless Texture Generation using Graph Cut, 大川 僚太, 吉田 大海, 飯國 洋二, 電子情報通信学会技術研究報告 = IEICE technical report : 信学技報,   2016 07 01
  • Automated GrowCut with Multilevel Seed Strength Value for Building Image, 上野 大, 吉田 大海, 飯國 洋二, システム制御情報学会研究発表講演会講演論文集,   2015 05 20
  • D-11-32 Binarization based on Conditional Image Enhancement and Character Stroke Width for Low-quality Document, Ikeda Shujiro, Yoshida Hiromi, Iiguni Youji, Proceedings of the IEICE General Conference,   2014 03 04
  • D-11-56 Two Step Inpainting Method Based on Space Distance and Energy Function of Pattern Similarity, Shibata Hiroki, Yoshida Hiromi, Iiguni Youji, Proceedings of the IEICE General Conference,   2014 03 04
  • D-11-36 EXTRACTION OF BLOOD VESSELS IN RETINAL FUNDUS IMAGES USING ADAPTIVE TOPHAT FOR NOISE REDUCTION, Ashida Ken, Yoshida Hiromi, Iiguni Youji, Proceedings of the IEICE General Conference,   2014 03 04
  • D-11-30 Estimation of a Waterline Using Spatio-Temporal Laplacian Filter for Draft Reading, Tsujii Takahiro, Yoshida Hiromi, Iiguni Youji, Proceedings of the IEICE General Conference,   2014 03 04

Misc

  • Description Image Inpainting Method based on Painting Texture, 吉田 大海, 画像ラボ, 29, 1, 24, 27,   2018 01 , http://ci.nii.ac.jp/naid/40021439304
  • Description of Scratch Inpainting Method based on Space Distance, 吉田 大海, 画像ラボ, 26, 11, 18, 23,   2015 11 , http://ci.nii.ac.jp/naid/40020641638
  • Description of a Conditional Image Enhancement for the Low-quality Document Image, 吉田 大海, 画像ラボ, 26, 4, 35, 40,   2015 04 , http://ci.nii.ac.jp/naid/40020420977