中口 俊哉

In this paper, we propose a method for improving image quality in order to match precisely the appearances of recorded video sequences with those of real scenes. Captured video and still images by cameras often have saturated and underexposed areas due to their limited dynamic range. Therefore, it is required to acquire images with a wide dynamic range and much tone data for the accurate reproduction of appearances. In addition, it is important to apply color reproduction processing based on human visual models to recorded videos because the appearances between the videos and real scenes differ due to the characteristics of human vision. In our method, a photometric high-dynamic-range (PHDR) video is generated by using a recorded video sequence and metadata of a camera such as the sensitivity (speed), aperture, and shutter speed. The PHDR video consists of the values of high-dynamic-range tone and absolute luminance. After the generation of a PHDR video, the video is processed by using color reproduction techniques based on spatiotemporal visual adaptation models for the accurate reproduction of appearances. Our results show that the proposed method reproduces fine appearance of videos which is the close appearance of real scenes. (C) 2007 Wiley Periodicals, Inc.

In recent years, many kinds of displays and recording systems for X-ray medical images have appeared. In this paper, sharpness of X-ray images displayed onto high-accuracy monochrome CRT and LCD monitors was analyzed and compared. Images recorded on the exclusive film and coated paper by inkjet printer and the wet-type and dry-type photo printers using a silver halide material were also analyzed and compared. The modified Gan's method was introduced to calculate the modulation transfer function (MTF) from the edge spread function (ESF). The results showed that the MTFs of the LCD, the inkjet image on the transparency, and the wet-type silver halide film image have fairly similar and good response in comparison with the inkjet image on the glossy paper and the dry-type silver halide film. It was also shown that the CRT has the worst response over the spatial frequency range. It was well correlated between the MTF and observer rating value. We thus consider the proposed method to be effective. (C) 2007 Wiley Periodicals, Inc.

腹腔鏡下手術において、先端部に2つのカメラを配置した立体腹腔鏡を用いてステレオ視を行い、臓器の形状を計測する手法がStoyanovらにより提案されている。Stoyanovらの手法では、画像全体を同時に対応付けて臓器の形状を復元するが、計算コストが高く、リアルタイム処理を実現できなかった。そこで本研究では、画像全体ではなく、画像の局所領域ごとに別々に対応付けを行うことで計算コストを削減し、よりリアルタイム処理に適したアルゴリズムを提案する。臓器模型を撮影した3種類の画像を用いて、計算コストおよび復元精度について、従来手法との比較を行った。その結果、対応付けの際に計算コストに強く依存する相関計算のテンプレートサイズを、従来手法の約16分の1とした場合でも、復元精度は4.68mm,7.18mm,7.44mmとなり、従来手法と同等の精度が得られた。(著者抄録)

In this paper, we propose a new projector-based display which can perform the color simulator for print industry. The proposed color simulator can change the color of print by projecting the image onto the print. A color of print can be matched to the desired color by projecting the image which is calculated to minimize the color difference between the colors of target print and current print. This current print is measured by digital camera or digital scanner. Ideally, spectral camera or scanner is expected to be used for accurate color simulation on the current print, but it costs a lot for practical application. Therefore, in this paper, we compared two methods for color matching, one is the tristimulus-based method with XYZ tristimulus values and the other is the spectral-based method with spectral values. As the result of computer simulation, the average color difference Delta E*(94) was 0.27 by the spectral-based method between the reflected radiance from the color of target print and the color of current print with projector, and the average color difference Delta E*(94) was 2.09 by the tristimulus-based method. The efficiency of the proposed system is verified by the subjective evaluation between the target and current print with appropriate image projection.

The purpose of this research is to propose an effective color metric which can predict the perceptual image quality for Retinex method. In this paper, we first give a brief introduction of three kinds of typical single Retinex methods to improve the color reproduction. And then, we state the process for obtaining the observer rating value from the subjective evaluation experiment performed under the sRGB illumination condition. Next, we introduce the S-CIELAB metric and propose a new metric on the basis of S-CIELAB metric that considers the gazing information. The average S-CIELAB color differences with and without the consideration of gazing information were calculated as the objective image quality measures. The correlations between the observer rating values and the objective image quality measures were calculated. The result shows that all of the average S-CIELAB color differences based on the gazing information are better correlated to the observer rating value than the average S-CIELAB color difference over the whole area. The average S-CIELAB color difference weighted by the gazing frequency over the gazing area shows the strong correlation with the observer rating value.

X線写真を用いた画像診断は古くから広く用いられてきた。近年では、ディジタルX線画像撮影システムが開発されたことにより、X線画像が銀塩写真のみならず、CRT、LCD、インクジェットプリンタといった様々な画像出力システムに表示・記録されるようになった。本論文では、画質を決める因子の中でも特に重要な鮮鋭度に注目し、湿式及び感熱式銀塩プリンタ、高精細モノクロCRT及びLCDモニタ、医用インクジェットプリンタのフィルム出力及び光沢紙出力の計6種類の医用画像出力システムの画質をそれぞれ解析評価し、画質比較を行った。Gans' methodを用いて、ナイフエッジ画像の出力応答を解析し、MTF(modulation transfer function)を算出することで鮮鋭度を評価した。Gans' methodにおけるシフトずれの問題について指摘し、独自に改良を加え測定精度を向上させた。評価の結果、インクジェットプリンタは銀塩プリンタに劣らない鮮鋭度を有しており、CRTは鮮鋭度において他のシステムより不利であることが分かった。また、主観評価実験を通して、本手法が鮮鋭度を評価する尺度として有効であることを示した。(著者抄録)

This paper presents a method of measuring a light field of a light source with high directional resolution using a mirrored ball and a pinhole camera. The light field describes a spatial and directional distribution of radiances from the light source. The directional distribution is expanded by a reflection on the mirrored ball, and the radiances are measured by a charge-coupled device (CCD) camera with a pinhole lens. The light source is laterally moved by a robot arm to measure the directionally expanded light field, and each pixel on a CCD can obtain the radiances from the light source through the pinhole lens with high directional resolution. The light field is estimated from the pixel value and the position of each pixel using a ray tracing technique. The light field of a krypton lamp was experimentally measured by the proposed method, and the accuracy of the measurement was evaluated against the irradiances measured by a spectro-radiometer at sample points. (c) 2006 The Optical Society of Japan

CT画像から再構築された上腕骨頭の三次元形状を左右間で比較し、欠損を定量的に評価する診断支援システムを構築した。三次元形状の再構成においては、新たに提案する判別基準法とヒストグラム解析を組み合わせた閾値決定法及び動的輪郭モデルにより、上腕骨の表面形状を抽出した。再構成した左右上腕骨の三次元形状を計算機上で重ね合わせ、直接比較することによって上腕骨頭の欠損部位を抽出できた。この際、両上腕骨頭の形状は鏡像対象であると仮定した。抽出された欠損部位の体積及び形状を解析したところ、本法の有効性を確認できた。

One of the most serious bottle-necks in fluid simulation occurs because the time step must be set to a small value to avoid numerical instability. To overcome this problem, this paper introduces a rigid core for the particles to maintain stability in particle-based fluid simulations. The fluid dynamics is calculated using the particle-based method with smoothed particle hydrodynamics (SPH) [Müller et al. 2003]. By using continuous collision detection (CCD) [Redon et al. 2001] among the rigid cores, unnatural compression and expansion can be prevented in large time step simulations. The calculation cost of CCD is negligible compared to that of small time step simulations, since CCD can be accelerated by a physics processing unit (PPU). Furthermore, this paper proposes a new real-time surface visualization technique that can also be assisted by a PPU. This technique generates large-scale high-resolution meshes compared to those generated by conventional lattice-based methods.

A spectral turn method is presented as a spectral based color correction method compatible with the standard RGB system by applying the idea of a multi-spectral method. The multi-spectral based color reproduction technique has been proposed for providing independent color of device and viewing conditions. However, it is difficult to replace the RGB devices with high cost multi-spectral devices since the latter have advanced only to the laboratory level. In the proposed method, the standard RGB values are converted into spectral information, then returned to RGB values for different viewing conditions. This process provides more accurate reproduction than the conventional forward process of the color reproduction. An experiment of psychophysical color matching was performed to evaluate the technique. The mixed chromatic adaptation model was applied to the method in the experiment, and the results showed the effectiveness of the proposed method. (c) 2006 The Optical Society of Japan.

We introduce a useful tool for controlling the skin melanin texture of facial photographs. Controlling the skin melanin texture is an important task in the reproduction of posters, TV commercials, movies, and so on. We used component maps of melanin, which were obtained by a previous method [J. Opt. Soc. Am. A 16, 2169 (1999)] as the first processing step. We propose to control the melanin texture continuously and physiologically, based on the analysis of 123 skin textures in our database. The physiological validity for the change of the melanin texture is confirmed by comparing the synthesized image with an ultraviolet image, which can be used to predict the change of melanin texture due to aging. The control processes are implemented on programmable graphics hardware, and real-time processing is achieved for a facial videostream. (c) 2006 Optical Society of America.

The system overlaying the stereoscopic images onto the haptic interface is proposed. In order to develop the tnaining systems based on virtual reality for medical doctors to experience medical treatments, it is important to improve the reality that visual and haptic information are synchronized. Our training system proposed in a previous paper for central venous injection uses the stereoscopic glasses and the haptic interface. This system has a palpation phase and an injection phase. During the injection phase, it is important that the user can see 3D injector images and the stylus of the haptic interface corresponding. In the proposed system, the user can look down the images projected from the display and reflected on the half mirror, and the haptic interface through the mirror. Using infrared LEDs wirelessly attached to the stereoscopic glasses and the infrared camera set under the display, the proposed system can show the 3D images considering with the face position in real-time. This system can achieve higher reality and more effectiveness than our conventional system. This system is evaluated by the experienced medical doctor as effective one.

人体切開手技トレーニングシステムを構築するため、切開時にメスに生じる力(切断力、粘性力、固定力)を再現する6自由度反力生成モデルを提案した。従来法は、反力計算においてメスの形状に制限がある、不均質な組織に対応していない、また3自由度の並進力のみの反力再現を行うため実際の操作感覚と差が大きいという問題が存在する。提案手法はこれらの問題を解決し、メスの形状と組織内部の粘性、硬性分布を考慮した反力計算モデルを用いた。本法は6自由度の力覚による反力提示が可能で、12名の非外科医による主観評価実験でも反力再現の有効性を確認できた。

腹腔鏡下手術の視野を通常の開腹手術の視野に近づけることによって、術者の負担を低減させることを最終的な目標とし、プロジェクターを用いた患者腹部への画像の投影による手術支援方式を提案した。まず腹腔鏡下手術の視野を通常の開腹手術の視野に近づける最初のステップとして、腹腔鏡画像を単純に患者の腹部に投影する方式(単純な投影方式)を提案した。腹部形状を取得することによって、投影による画像の歪みを補正し、術者の視点位置をリアルタイムで取得することによって、投影画像を術者の視点移動に対応させた。通常の手術方式との比較実験により鉗子挿入操作の有効性を確認した。

偏角分光イメージング法は物体の形状や色, 光沢情報を正確に記録することができ, ディジタルアーカイブへの応用が期待されている. 一般に偏角分光イメージング法ではカメラ・照明の配置を多数変化させるため, 計測時間が膨大となる. 本提案手法では計測時に表面粗さパラメータを仮定し評価関数を構築する. 遺伝的アルゴリズムにより最適化した配置を用いた撮影を行い, 計測値の妥当性を検証する. こうして物体表面の粗さを画素単位で計測する事により, 複雑な表面粗さ分布を持つ物体に対しても適応可能な, 偏角反射特性計測のための効率的なカメラ・照明配置の決定方法を提案する. 本手法の有効性を計算機シミュレーションにより評価し, また自動制御撮影システムを構築して画像再現実験を行った.

In conventional laparoscopic surgery, one medical doctor who only holds the laparoscope during the surgery is required. Therefore, in this study, we introduce an automatic tracking and zooming system for laparoscopic surgery by means of a steady laparoscope on the abdominal surface. The system can be added to a conventional laparoscopic system and can analyze camera images in real-time. The markers of this system are attached at the top of the forceps and extracted, and the focusing area can be estimated. In this study, we introduce two marker tracking algorithms; one tracks a single marker on the forceps, which is assumed as the focal point; and the other tracks two different colored markers, which can estimate a more practical focusing point. In order to perform high-quality real-time movie processing, the system uses a CPU and GPU (graphics processing unit) in parallel. The tracking algorithm is carried out by the CPU, image enlargement is carried out by the GPU, and then a high frame rate and low delay time are achieved. For the automatic smooth camera work, a new control model with many parameters is proposed. Through animal experimentation was performed six times, and appropriate values of parameters evaluated by two experienced doctors were obtained. After the evaluations of the experienced doctors, the effectiveness of the system was proven: cholecystectomy successfully conducted in animal experiments without the support of a scope-holding doctor.

This paper presents a novel class of cellular neural networks (CNNs), where output of a cell in the CNN is given by the piecewise-linear (PWL) function having multiple constant regions or a quantization function. CNNs with one of these output functions allow us to extend CNNs to image processing with multiple gray levels. Since each cell of the original CNN has the PWL output function with two saturation regions, the image-processing tasks are mainly developed for black and white output images. Hence, the proposed architecture will extend the promising nature of the CNN further. Moreover, the hysteresis characteristics are introduced for these functions, which make tolerance to a noise robust. It is demonstrated mathematically that tinder a mild assumption, the stability of the CNN which has an output function with hysteresis characteristics is guaranteed, and the impressive simulation results are also presented.

We propose hysteresis neural network solving combinatorial optimization problems, Box Puzzling Problem. Hysteresis neural network searches solutions of the problem with nonlinear dynamics. The output vector becomes stable only when it corresponds with a solution. This system does never become stable without satisfying constraints of the problem. After estimating hardware calculating time, we obtain that numerical calculating time increases extremely comparing with hardware time as problem's scale increases. However the system has possibility of limit cycle. Though it is very hard to remove limit cycle completely, we propose some methods to remove this phenomenon.

We propose a hysteresis neural network system solving NP-Hard optimization problems, the N-Queens Problem. The continuous system with binary outputs searches a solution of the problem without energy function. The output vector corresponds to a complete solution when the output vector becomes stable. That is, this system does never become stable without satisfying the constraints of the problem. Though it is very hard to remove limit cycle completely from this system, we can propose a new method to reduce the possibility of limit cycle by controlling time constants.