-
Kazuya Miyakawa, Keishi Nishikawa, Jun Ohya, Kenji Hashimoto, Atsuo Ta ...
Session ID: S1-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this paper, in the case where the three-dimensional coordinates of the end points of the cross-point of the vertical ladder, which is necessary for raising and lowering the disaster-responsive robot WAREC-1 attached to the vertical ladder, are unknown, the three-dimensional point cloud data acquired by LRF is analyzed, We proposed a method to estimate three-dimensional coordinates required for WAREC-1 to raise and lower a vertical ladder, and investigated. In the experiment, we investigated the installation position and angle of the sensor, and improved the estimation accuracy by optimizing the installation position and angle of the sensor. Experiments were conducted on vertical ladders composed of cylindrical / prismatic members, and the average values of the estimation errors were less than 7.1 mm and 9.1 mm, respectively. The proposed method can estimate the three dimensional coordinates of the cross-point required for the leg type robot WAREC-1 to raise and lower the vertical ladder.
View full abstract
-
Yukihito KUNITAKE, Munetoshi IWAKIRI, Tsuneki SAKAKIBARA, Kan-ichiro M ...
Session ID: S1-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Many structures were collapsed by The 2018 Kumamoto earthquake. When the earthquake occurred, we can use the aerial images effectively to grasp the situation of damaged buildings in wide area. But people need long time to know the damage situation because they classify the collapsed building by them eyes.
In this paper, we approached the automated method to detect the collapsed structures fast and correctively from the wide aerial images, and studied the capacity of the method. We detected the collapsed buildings from the aerial images immediately after The 2016 Kumamoto earthquake to use the method, could get the number of the clearly collapse buildings.
View full abstract
-
Junya MORIMOTO, Takuya HAYASHI, Hiroyuki FUJIMOTO, Hiroyuki ISHI, Jun ...
Session ID: S1-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In recent years, we are aware of many environmental problems, such as environmental destruction,
reduction of wild animals. It is required to make use of autonomous robots for environmental monitoring to
predict disasters and grasp situations.
In this research, we consider image processing technology that contributes to environmental monitoring by
WAMOT (Waseda Animal Monitoring Robot), an environmental monitoring robot developed for creating
environment and animal ecology maps. WAMOT uses a charging station to withstand long-term autonomous
observation. When the battery of the robot runs short, it automatically returns to the charging station. As
soon as charging is completed, repeat the operation of starting the observation again to grasp the
surrounding environment. In this research, we aim to make the Path Planning Based on SLAM and
Semantic Segmentation by Deep Learning.
We propose a method the path planning that based on SLAM and cost map reflecting the result of Semantic
Segmentation. SLAM in irregular areas like forests is estimating the robot's state from the stable self
position estimation and combining wheel odometry. Cost map based on the environment map (self position
estimation result) and the estimation result by the deep learning from RGB image. As a result of the
experiment, path planning from cost map generation reflecting the result of environment estimation could
be performed.
View full abstract
-
Takuya OKITA, Hiroki KOJIMA, Sho OOI, Mutsuo SANO
Session ID: S1-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
There are 310 thousand the visually impaired people in Japan. They use a white cane during walking outside the home. However, the white cane is difficult to get surrounding information (Ex. Traffic light, Signboard, etc...).
The aim of the research is to develop a system that the impaired people are able to walk safely. Specifically, this research uses an egocentric vision and is able to recognize “traffic light”, “Signboard”, etc... using deep learning in real time. Therefore, the system is able to estimate dangerous situation, feedback the result to the user.
View full abstract
-
Kunio OHNO
Session ID: T-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Rapid technical advances and globalization in the IT field call for liberal artistic skills that allow changes in the market to deal with all rounds, which can follow change, rather than to break down specialized stress.To that end, we will endeavor to flexibly expand the domain of expertise, strive to acquire skills in areas that we were not good at, and we need to strive to grasp and acquire universal values and to direct our way of life.
View full abstract
-
Norikatsu FUJITA , Kazushige MATSUMOTO, Masahiro YOKOYAMA, Hideyo TSUK ...
Session ID: T-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
-Design of skill assessment in problem behavior scenes-
Sho AOKI, Norikatsu FUJITA, Hiroshi TAKESHITA , Mahoro ISHIHARA, Ribun ...
Session ID: T-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
[in Japanese], [in Japanese]
Session ID: T-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
-A Case Study of a Student Entrepreneurs in Sendai-
[in Japanese]
Session ID: T-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Kunio OHNO, Toshio KAJIWARA
Session ID: T-6
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We will introduce XML as a standard format for voluntarily creating and proposing virtual museum content relating to geomagnetic reversal stratification understanding and how to use the Lisp language to simulate physical phenomena
View full abstract
-
Tomo NODA, Nonoho TANIGUCHI, Kitahiro KANEDA, Keiichi IWAMURA, Nobuyu ...
Session ID: S2-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this paper, we propose a genuineness judgment of a bar code printed using an ink jet printer. In recent years, barcodes have been widely used for drug labels or aircraft boarding passes, but counterfeiting of barcodes has become a major problem. In order to solve this problem, from our experiments, it is observed that the shape, size and location of printed dot are randomly distributed even in the same gray level. We consider to utilize this characteristics to separate genuine print and counterfeited print. In this study we propose the novel approach to detect counterfeited 2D symbol or 2D code generated by inkjet printer with image preprocessing and Locally Likely Arrangement Hashing scheme. As a result, 55% of the registered barcode could be correctly identified.
View full abstract
-
Chisato OKAMOTO, Masahiro ISHIKAWA, Emi HASHIMOTO, Kazuma SHINODA, Mad ...
Session ID: S2-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Hiroki Satake, Shutarou Kodama, Katsutoshi Inomoto, Kanami Ikeda, Kats ...
Session ID: S2-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Aya KUBOTA, Chinatsu MORI, Seiichi GOHSHI
Session ID: S2-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Imaging devices always cause noise. Noise reducer (NR) is one way to reduce noise in videos. Although the conventional NR reduce noise, it decrease resolution of videos. Especially, a recursive type NR generates degradation such as motion areas blur and trailing. One of authors proposed a non-recursive type NR algorithm; however, the proposed NR has not compared with the conventional one with a quantitative quality test. In this study, a subjective assessment was conducted to prove the ability of the non-recursive type NR. The results show that the proposed NR algorithm is good for resolution-preserving, and noise-reducing.
View full abstract
-
Naoya SAKUMA, Kazuhiro MIYAMAE, Seiichi GOHSHI
Session ID: S2-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Due to advances of imaging devices, prices of video cameras have become reasonable. Although cost of video shooting/editing was expensive, it has also become reasonable prices. It makes cut the cost of security cameras and the market of it is expanding every year. In contrast even though the advances, noise is an essential issue of videos. It is conspicuous when the videos are shot under poor lighting conditions. Nose degrades the video quality. Although it is impossible to remove noise completely, noise reducer, a signal processing, can reduce the noise in the video. However, it is necessary to detect the accurate noise levels in the video for the noise reducer to work effectively. In this paper we propose a novel method to detect the accurate noise level in general videos.
View full abstract
-
Ayumu WADA, Kyosuke SHIMADA, Seiichi GOHSHI
Session ID: S2-6
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
It is impossible to introduce steady characteristics for images/videos even though stochastic process or statistics are applied. The examples of the steady characteristics are the Gaussian noise or the same luminance level on the whole screen. There are many methods to analyze the steady information sources and the analysis results accord to the real images/videos. In contrast it is well known that local areas in images/videos have similar characteristics. In this paper we focus on the local characteristics of images/videos. Using this characteristics we propose the novel method to convert general images to the pseudo-steady information sources. The applications of the proposed method is also discussed as a coding technology.
View full abstract
-
Piyarat SILAPASUPHAKORNWONG, Masahiro SUZUKI, Hideyuki TORII, Kazutake ...
Session ID: R1-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We have been studying a method of information hiding inside an object fabricated with a resin material using a 3-D printer by forming pattern inside the object that expresses information. In this report, we report the new method of forming an internal pattern with a near infrared fluorescent resin material in order to improve the accuracy of reading embedded information. In this method, when an object is irradiated with near-infrared light of a specific wavelength that excites the fluorescent resin material, the fluorescent resin emits light at a wavelength different from that of the excitation light. Therefore, if only the light emitted by the fluorescent material is used for capturing near infrared image by using the optical filter, only the image of the pattern embedded in the object can be captured, and information reading with high accuracy can be expected. Using a fused deposition-modeling 3-D printer with two nozzles, the main body of an abject was formed with one nozzle using ABS resin, and the other nozzle formed an internal pattern with ABS resin containing fluorescent dye. We confirmed that clear images of internal patterns could be captured and the effectiveness of this method was demonstrated.
View full abstract
-
Hiroshi UNNO, Kazutake UEHIRA
Session ID: R1-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We propose a digital signage display technique that can invisibly embed a binary meta-image in a color component of a displayed cover-image. The technique temporally and spatially modulates intensity in a color on the cover-image according to the binary meta-image. We also propose a method to extract the meta-image in a displayed cover-image from captured image using video cameras. We first examined invisibility in terms of colors for a temporally and spatially intensity modulated meta-image in a displayed cover image. We also verified that invisibly embedded binary meta-image could be extracted from captured image by using MPEG4 AVC video format.
View full abstract
-
[in Japanese]
Session ID: Q-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Takashi HOSONO, Shuhei TARASHIMA, Jun SHIMAMURA, Tetsuya KINEBUCHI
Session ID: R1-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Generating instance candidate regions of arbitrary objects in an image is called object proposal is important in various areas of computer vision. We propose an accurate and fast object proposal method using depth images. Existing object proposal methods can be roughly divided into two categories: window scoring and object region extraction. Of the two, window scoring methods usually have higher efficiency but often suffer from the effects of object texture. In addition, these methods estimate regions that capture multiple objects. We tackled this problem by defining a novel objectness measure from depth images. Since the depth edge of an object is often a closed loop, the proposed method evaluates objectness on the basis of the assumption that regions containing an object should have high depth edge density at the outer region and low depth edge density at inner region. We also took the edge density uniformity at the outer region into consideration. Using our objectness measure, the proposed method estimates individual object regions accurately even in cases where existing methods estimate regions that capture multiple objects. Experimental evaluations demonstrated that our method can estimate proposals faster and more accurately than state-of-the-art methods on a challenging dataset of complex crowded scenes.
View full abstract
-
Yongqing Sun, Shingo Ando, Jun Shimamura, Tetsuya Kinebuchi
Session ID: R1-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
On this report, we would be presenting our observation on various object detection methods on manhole dataset. The first method that we tried is the original Single Shot Detector built on top of Caffe; we tried both input size available 300x300 and 512x512. For both experiments, we got pretty promising results. In addition to those experiments, we also tried Single Shot Detector built on Keras and did basically the same experiment. The result is not that great in comparison to SSD on Caffe.
View full abstract
-
Akira ASADA, Kohei TAKEUCHI , So SUZUKI, Makoto J. HIRAYAMA
Session ID: R1-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
As a safety measure when Head-Mounted-Display is worn, we propose a system for BLE communication. This system uses smart phone as central BLE device. It measures the distance by receiving smart phone from peripheral BLE devices to BLE communication. And it issues a warning according to the distance.
View full abstract
-
Masafumi KOSUGI, Keisuke UTSU, Sachi TAJIMA, Makoto TOMITA, Yoshitaka ...
Session ID: R1-6
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In order to minimize the damage at the time of a disaster, it is important to collect and distribute information quickly and accurately. Therefore, Twitter, which is characterized by high immediacy and many users, is attracting a great deal of attention for its use in disasters. Based on this background, we have implemented a disaster information tweeting and mapping system DITS and DIMS in previous research and have improved it. In this paper, we report improvements of the system based on user's opinion and extension to the peacetime utilization system (local information tweeting and mapping system LITS and LIMS).
View full abstract
-
Naoto KAWAMURA
Session ID: R2-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
I propose an invisible digital watermarking which enable to embed two types of information. The main information consists of bit data of the character code representing Green-Noise characteristics, and sub information consists of bi-level pattern of logos and figures representing checkerboard pattern. These patterns are combined with image data on spatial domain.
The Green-Noise pattern is difficult to recognize visually and shows robustness against various attacks. The spectrum of sub information is distributed in a higher frequency region which does not overlap with the spectrum of Green-Noise pattern, resulting in a fragile watermark. These two types’ watermarks can be extracted by different watermark extractors. It can be applied to protection of copyright protection, detection of tampered position, tracking of illegal outflow etc. Further, the sub information can be used as a synchronization signal for extracting the watermark of the main information or as a copy-forgery-inhibited pattern for printing.
View full abstract
-
-Fake Technologies in Detective Dramas-
Seiichi Gohshi
Session ID: R2-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In detective dramas and films, police have special technologies to improve security image quality.
Although there is not such kind of technologies and they are completely fakes, many people misunderstand they exist. In this presentation this misunderstanding would be solved and the limitation of image quality improvement technologies will be discussed.
View full abstract
-
Daisuke KOBAYASHI, Ken NAKAMURA, Hiroe IWASAKI, Atsushi SHIMIZU
Session ID: R2-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We have developed a new real-time HEVC encoder for multi-bitrate streaming. High-quality video encoding and bandwidth reduction ware realized by dynamic bitrate control while keeping image quality.
View full abstract
-
Mei KODAMA
Session ID: R2-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this study, we consider a video management method to reduce the cache efficiency in a video distribution system with cached data. In particular, we consider the effect of decreasing the estimation error in the priority estimation method in each used video and the influence on the cache efficiency by using the variable model with the access number and the usage quality. In addition, we consider the relationship between the estimation error and the cache efficiency when the length of the reference interval period is changed.
View full abstract
-
Takashi NISHIHAMA, Kohei TOKOI
Session ID: P-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this paper, we report a method to interactively perform hair simulation by interaction between water
and hair. The proposed method approximates the hair as a set of mass points and line segments and
simulates the motion of the mass point by using the spring damper model, thereby reproducing the shape
change of the hair. In addition, a simulation method based on the lattice method is used to reproduce fluid
motion and hair wetting effect that affect hair movement. Through these methods, the proposed method
can reproduce changes in hair shape and shadow due to water adhesion to hair in real time.
View full abstract
-
Isseki MIWA, Shigeru AKAMATSU
Session ID: P-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this study, we used Kinect V2 RGB-Depth sensor to measure three-dimensional shape of faces and calculated the information of the differences among several individual. Based on Principle Component Analysis on the obtained information, we tried to verify the possibility of generating various expressions from an inexpressive face of an arbitrary individual.
View full abstract
-
Shutarou Kodama, Hiroki Satake, Hiroyuki Kowa, Takashi Hoshiba, Eriko ...
Session ID: P-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Hiroki KIYOZUMI, Taichi EZAKI, Masahiro ISHIKAWA, Naoki KOBAYASHI
Session ID: P-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We study about pulse wave measurement system using video image of patient for telecare consultation. For getting the precise shape of pulse wave, we focused on a method using multi-band spectral images i.e. with RGB images and one band near infrared (NIR) image.
The experimental result shows that component of independent component analysis of signals from green, red and NIR images was most similar shape of plethysmogram of finger. And the component using area where the variance of signal of NIR had been maximum was most effective.
View full abstract
-
Hirotaka YAMASHITA, Ayumi MATUURA, Naoki KOBAYASHI, Masahiro ISHIKAWA
Session ID: P-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
To clarify the relationships between visual induced motion sickness (VIMS) and visual environments, we focused relationships between VIMS and ambient illuminance. The experimental results show that there is no significant difference between subjective scores in a bright room condition and in a dark one. Concerning to LF/HF ratio, there are some difference of signals between these two conditions but no significant.
View full abstract
-
Ayako KATOH, Yuki UMEMURA, Yasuhiro FUKUI
Session ID: R3-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Kunihiko TAKANO, Syuntaro TOMINAGA, Tatsuya IKEUCHI, Koki SATO, Kikuo ...
Session ID: R3-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this papaer, we descrive about the projecting technique of the holographic 3D imgaes with the use of the fog screen. As a result, we found that our method proveds effective for holographic image projection with high stability.
View full abstract
-
Kunihiko TAKANO, Minoru ASUMA, Kakeru IDA, Koki SATO, Kikuo ASAI
Session ID: R3-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this paper, we describe about the characteristics of holographic reconstructed images with the use of blue-violet laser light source. As a result, we found that our method suggests the possibility of good results to color holographic image reconstruction.
View full abstract
-
Junichi HARA, Takaaki ISHIKAWA, Osamu WATANABE, Shigetaka OGAWA
Session ID: OS1-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
More than 25 years ago, the original JPEG standard was developed and published. After the historical success of the International Standard, various International Standards for still image compression have been deliberated, and new standards are still under discussion. In addition, revisions of previously developed International Standards, standardization of next generation still image compression, and application of blockchain technology are also being studied by JPEG committee.
In this paper, we outline these standard techniques. First, we describe the standard development process necessary for grasping the development stages of each JPEG family standard, and we compare the technical features of the standardized image coding methods. After that, we express the position of the standards currently under development such as JPEG 2000 Part 15 (HTJ2K), JPEG XS, JPEG Systems (JPEG Privacy & Security, JUMBF, JPEG 360), and JPEG Pleno. We also address the outline of each technology. In addition, we also touch on the latest deliberation status including the next-generation image compression and blockchain application currently being considered for standard development.
View full abstract
-
- 4K/60P HEVC Encoder LSI(NARA)-
Hiroe IWASAKI
Session ID: OS1-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Kazuma SHINODA, Takumi NISHIOKA, Naoki KOBAYASHI, Madoka HASEGAWA
Session ID: OS1-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
A region of interest (ROI) coding function has been defined for JPEG2000. ROI coding is a technique for improving a partial image quality by assigning priority to ROI. However, when multiple ROIs need to be coded, the coding performance degrades since some non-ROI coefficients are encoded. We improve the wavelet transform and EBCOT in JPEG2000, and show that the proposed method achieves higher PSNR than JPEG2000 ROI coding.
View full abstract
-
Kazuya SATO, Kazuto KAMIKURA
Session ID: OS1-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Recently, a hyperspectral camera has got attention as a new sensing device. In a remote sensing field, a hyperspectral camera is often loaded on small machines with limited power consumption to measure the data. Therefore, low encoder complexity is required. In this paper, we focus on Distributed Source Coding (DSC) which is a compression technique that realizes low encoder complexity. We apply coefficients of the Weighted Prediction (WP) technique to Side Information (SI) in order to improve the quality of SI. Furthermore, we construct compression coding system based on DSC, and aim to improve the compression efficiency. We estimate quality of SI by the proposed method by comparing with simple interpolation or extrapolation methods.
View full abstract
-
Jaime SANDOVAL, Kazuma UENISHI, Munetoshi IWAKIRI, Kiyoshi TANAKA
Session ID: OS1-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
One of the most common types of parametric surfaces found in both industrial and urban environments are spherical. While more efficient and precise algorithms are needed for practical applications, conventional methods still are slow or require additional surface information, therefore, we proposed a sliding voxel based algorithm to locally fit spheres into point clouds and casting votes into a sparse 4D Hough accumulator.
View full abstract
-
Tomoyuki NISHITA
Session ID: Q-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Masahiro Yamaguchi
Session ID: OS3-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
This paper overviews the recent trend of spectral imaging technology, and introduces the technical report “multispectral image format” prepared by CIE TC8-07. Then the future issues on multispectral and hyperspectral technologies are discussed
View full abstract
-
Daisuke MIYAZAKI, Ikumi NAKANE, Shinsaku HIURA
Session ID: OS3-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Metamerism is the phenomenon by which two objects are recognized as having different colors under one light source and the same color under another light source. In this article, the authors propose a method for creating trick artwork using metamerism. In order to achieve the goal, we have constructed the multi-spectral database of LEDs and oil paints. Intended illuminant can be designed by a mixture of LEDs, and also, intended paints can be designed by a mixture of oil paints. In this work, illuminants and paints are designed to achieve metamerism such that two oil paints used in a piece of artwork look the same under one light but different under another light. The experimental results show that metamerism is generated between the two light sources and the two object colors.
View full abstract
-
Norihiro TANAKA, Seungwan HONG, Kiyoshi TANAKA
Session ID: OS3-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
This paper propose a method for estimating scene illumination for multispectral rendering by using RGB color camera. The spectral distribution of the light source of the scene is estimated as an omnidirectional spectral image for IBL assuming that the spectral distribution of the illumination light source is the spectral distribution of black body radiation. Finally, in order to confirm the effectiveness of the proposed method, CG of human skin was reproduced under the scene illumination environment measured by the proposed method.
View full abstract
-
Kazuma SHINODA, Madoka HASEGAWA
Session ID: OS3-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We propose a novel demosaicking method for Fourier specral filter array (FSFA). The proposed method consists of 2 steps: spatial and spectral reconstruction. In the spatial reconstruction, we separately estimate low- and high-frequency components decomposed by a discrete wavelet transform. The low-frequency components are estimated by the bilinear interpolation, and the high-frequency ones are done by compressive sensing because of their sparsity. Experiment results shows that the proposed methods outperforms other existing demosaickings especially in the spatial reconstruction.
View full abstract
-
Yohei TAKARA
Session ID: OS3-5
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Hyperspectral imaging (HSI) is a growing imaging technology originated from remote sensing that acquires spatially resolved spectra from an object. Recent advances in image sensor technologies and computational power enables the development of small hyperspectral camera systems and HSI has been applied in diverse researches and industrial areas. We describe a brief overview of HSI and introduce some cases of hyperspectral applications and future experience with spectral imaging.
View full abstract
-
- Aiming to realize a ultimate comfortable space -
Kitahiro KANEDA, Emiko SAITO, Hiroyuki OKADA
Session ID: OS2-1
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Report of a social experiment for a collaboration of music and visual presentation.
View full abstract
-
[in Japanese], [in Japanese], [in Japanese], [in Japanese]
Session ID: OS2-2
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
[in Japanese]
Session ID: OS2-3
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
-
Akinori TANI, Yuichiro YAMABE, Hirokazu KOBAYASHI
Session ID: OS2-4
Published: 2018
Released on J-STAGE: January 10, 2022
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
In this paper, results on optimal location of security cameras around an elementary school using Genetic Algorithm are presented and discussed.
View full abstract