In recent years, it is said that communication speed of next-generation cellular phones will reach 100Mbps. Accordingly, demand for robust error correction codes with high-speed processing is increasing. High-dimensional discrete torus knot code (torus code) performs well in fields with many errors. We propose a new and unique synchronization system that uses the torus code. It performs error correction and synchronization processing effectively and simultaneously. According to results of comparison with a conventional system (HDLC:High Level Data Link Control), performance (Loss of Lock Probability) of the proposed synchronization system is as much as 5 figures better than the conventional HDLC in the 10^-4 field (Input BER). By applying a LSI technology, we have developed a FPGA(3Dm5, 125bit, r=0.512) with the proposed synchronization system and have built a high-speed MPEG communication device which uses the newly developed FPGA. The high-speed MPEG communication device can transmit a video signal of 20Mbps.

抄録全体を表示

虹彩パターンは指紋や顔に比べて安全かつ高精度な認証が可能であることから生体認証要素として理想的であるが，一般に充分普及しているとはいえない．その原因として挙げられるのは，既存の虹彩認証システムが複雑な装置を用いることによる導入の難しさや，至近距離での入力画像の撮影を求めることによるユーザビリティの低さである．撮影された画像から個人の特徴を記述するには，正確な虹彩領域分離が不可欠である．しかし小型でシンプルな撮影装置を用いつつ撮影距離の制約を排除することを目標とすると，既存手法が前提とする瞳孔の絶対的明度が不安定となるために虹彩領域分離の困難さが問題となる．そこで本稿では，そのような条件でも安定な虹彩領域分離を可能にする手法を提案する．提案手法に代表的な特徴抽出ならびに照合手法を組み合わせて実施した評価実験から示されたのは，提案手法が従来対応困難な画像においても安定な虹彩領域分離を実現すること，そして既存手法に比べて全面的により優れた認証性能を有することである．また，新規に提案したノイズ除去手法を併用することで，虹彩領域の輪郭が曖昧な場合でも高精度な認証が可能となることも示されている．

抄録全体を表示

KARAOKE is a popular amusement for old and young. Many KARAOKE machines have singing evaluation function. However, it is often said that the scores given by KARAOKE machines do not match human evaluation. In this paper a KARAOKE scoring method strongly correlated with human evaluation is proposed. This paper proposes a way to evaluate songs based on the distance between singing pitch and musical scale, employing a vibrato extraction method based on template matching of spectrum. The results show that correlation coefficients between scores given by the proposed system and human evaluation are -0.76∼-0.89.

抄録全体を表示

This paper proposes a pitch estimation method suitable for singing evaluation incorporable in KARAOKE machines. Professional singers and musicians have sharp hearing for music and singing voice. They recognize that singer's voice pitch is “a little off key” or “be in tune”. In the same way, the pitch estimation method that has high frequency resolution is necessary in order to evaluate singing. This paper proposes a pitch estimation method with high frequency resolution utilizing harmonic characteristic of autocorrelation function. The proposed method can estimate a fundamental frequency in the range 50 ∼ 1700[Hz] with resolution less than 3.6 cents in light processing.

抄録全体を表示

本稿では，指紋画像の特徴を有限なパタンの組合せに近似し，代替的な指標の系列として表す手法を提案する．本手法はあらゆる人の指紋画像の特徴を単一の枠組みで効率的に記述可能とするもので，個人を特定しうる情報を保持せず特徴を記述するため，生体情報を用いた個人認証における実用性も高い．指紋画像の特徴を直接照合する場合と比べた評価から，本手法は認証精度を損なうことなくより高速な照合が可能であることが示された．

抄録全体を表示

In this paper, we propose the way to modify gradient of fingerprint images. We modify the gradient using fingerprint area and contour, without extracting end points or bifurcation points called “minutia" from fingerprint images. It takes 3.2[ms] to modify fingerprint gradient by using a personal computer which operates at 1000[MHz] clock frequency. Our program code is so light weight that it is able to be embedded in fingerprint verification devices which operate at lower speeds. We evaluated our algorithm by using fingerprint images tilted respectively 0°, 5°, 10° and 15°. Without modifying gradient, we achieved 90.8[%] recognition rate by images tilted from 0° to 5°, while false acceptance rate is 0.01[%]. By using our gradient modification algorithm, recognition rate raised from 90.8[%] to 98.2[%]. On the other hand, 19.4[%] recognition rate was achieved by images tilted from 10° to 15° without gradient modification, and the rate raised to 89.2[%] with gradient modification. The users who don't get used to fingerprint devices tend to input his finger tilted on scanner. Thus, it is expected that the fingerprint verirification device will be improved to be more friendly for all peoples, by using our gradient modification algorithm.

抄録全体を表示

Recently, small sized area type fingerprint sensors have been installed into popular mobile devices for fingerprint verification. Because of high-density integration and convenience, small area-type sensors have been chosen for biometric authentication on many mobile devices. However small area-type sensors capture small area fingerprint images and cause lack of minutia points on minutia verification. We have developed a new matching algorithm available to small area type fingerprint sensor. In the proposed method, polar coordinates sampling is used in extracting feature and DP matching is used for matching features. The results of the experiments show that proposed method performed high accuracy verification using a small area-type sensor.

抄録全体を表示

In these years, demand for three-dimensional measurement technology is higher. However, in many of the traditional three-dimensional measurement methods, such as the method using phase shift patterns projection, it is difficult to measure shape of specular reflective and transparent objects, because power of reflected light is too high from specular reflective objects or too low from transparent objects, so that information of reflected light cannot be acquired accurately by digital camera. Digital holography has been used for resolving this problem. In traditional digital holography, interference fringe of measure object's light wave and reference light wave is recorded by digital camera. Using Fresnel diffraction principle, phase information of measure object's light wave can be required and shape of measure object can be calculated using interferometry principle. It can be used for reflective and transparent objects measurement, but as the principle of Fresnel diffraction, DC and conjugate components exist in calculated phase information as noises, so accuracy of measurement is lower. Phase shift digital holography is able to remove noises using special devices to change phase of reference light wave, but the devices are high price and the system is complex, so practical application of phase shift digital holography is difficult. In this paper, we propose a new method of phase shift digital holography using light detection principle. Simultaneously, we propose a digital holography reconstruction method using the convolution approach rather than the traditional Fresnel approach. The proposal method realized a compact phase shifting holography system with low price and a high accuracy three-dimensional measurement method. Therefore, we are able to realize the practical applications of digital holography with low price and higher accuracy.

抄録全体を表示

In order to segment robustly character areas of serial numbers in weak-contrast and noisy license plate images which are extracted from the front view of cars by the method described in our previous paper, we propose a useful method which is constructed by artificial neural networks corresponding to each digit of the serial numbers. The ability of proposed method is compared to that of a histogram-based method which is used as an example of conventional methods. In the histogram-based method, touching characters and noisy characters are often mis-segmented. The other hand, the network-based method can segment precisely even for those characters. For 595 license plate images, a segmentation rate of 96.7% in the histogram-based method and a rate of 100% in the network-based method with an additional second search are gained. Then, a network-based recognition method for the segmented characters is investigated. By randomly varying the segmentation size of trained characters in learning process of the network, all the serial numbers in the 595 images can be recognized.

抄録全体を表示

In surface mount technology, many kinds of image processing methods are used in inspection and positioning for mounting surface mount devices (SMDs) onto PCBs (Printed-Circuit Boards). Currently, such methods require part shape data to detect and inspect the position of parts. These shape data are generated by manual operations which require costs (e.g. operators, times). Automation of the shape data generation has been performed by hierarchical classification method proposed in previous research. Our research focuses on SMDs of complex shape, which was not supported in previous research. We proposed a lead detection method combining frequency analysis and AdaBoost to generate more accurate part shape data. We evaluated our proposed method with 70 parts that were not supported by previous research. The result of evaluation showed that the proposed method reduced false detection rate of leads detection from 24.5[%] to 2.4[%].

抄録全体を表示

This paper presents a fingerprint verification method based on the linear predictive coding (LPC) analysis. This method is useful for realizing on small-size hardware because the computation to extract spectrum patterns from fingerprint images consists of one-dimensional signal processings. In evaluation, the recognition rates of the methods using the LPC spectrum, LPC cepstrum, and the group delay spectrum (GDS) are compared. The results showed that the highest recognition rate is obtained for the GDS. The true acceptance rate was 97.4% under true rejection rate of 100.0%. The true acceptance rate improved to 98.5% by the DP matching with free start and end points. Furthermore, we showed that 2-bits quantization of the GDS brings no decrease of the recognition rate.

抄録全体を表示

近年，上場会社において，不適切な会計処理の発覚により，過去数期分の有価証券報告書の訂正とともに訂正内部統制報告書を提出している事例が多く見受けられる。不適切な会計処理が発覚した場合には，経営者による内部統制の評価が適切であったかどうかが問題となり，また，監査人は当初の財務諸表監査および内部統制監査で表明した監査意見が適切であったかどうかが問題となる。内部統制報告制度の導入後における開示義務違反に係る告発・課徴金納付命令事案を対象として，処分対象期の内部統制の評価結果および監査人の対応について分析し，内部統制報告制度の問題点および今後の課題を提言している。本稿において明らかとなった実態は，経営者による内部統制の評価・監査人による内部統制監査の簡素化や，新規上場企業の内部統制監査の免除といった負担軽減の方向に一辺倒の制度設計に警鐘を鳴らしているように見える。

抄録全体を表示

In this paper, an effective fingerprint verification method for varying qualities of fingerprint images is described. When an optical sensor is used for imaging a fingerprint, the image quality is degraded especially due to a lack of moisture on the finger surface. Under such a condition, the captured fingerprint images are often faded. First, we investigated a method for improving the quality of faded fingerprint images, which is based on high frequency emphasis of the image using Fourier transformation. After the image is enhanced, the peaks of the group delay spectra (GDS) which are the features extracted for verification are more distinct than those extracted from the original image. When the reference patterns are selected from the fingerprint data obtained in April, the verification rate with the high frequency emphasis is 90.8%, while the rate without the emphasis is 75.3%. However, for the data obtained in February which include many faded fingerprints, the verification rate strongly decreases. Second, we examined a method to adapt differences in seasonally dependent image quality based on updating a reference pattern in terms of the accepted input pattern. Using this method, when the reference patterns are composed of the February data, the verification rate is improved from 57.1% to 85.2%. Furthermore, by resetting the verification threshold value whenever a reference pattern is updated, the rate increases to 96.5%, which is equivalent to that when the all seasonal data are registered.

抄録全体を表示

In the existing methods of a moving object detection using image recognition technology, they have processed an obtained whole image, and have detected a moving object in a picture. However, when a moving portion in the scene is hidden by some obstacles, the recognition of a moving object is sometimes difficult. In this study, we propose a novel method of discriminating moving objects, such as a person or a vehicle. We use only a narrow and tall area of the video called the Strip Frame Image for detecting moving portions. By using this method, we are able to obtain the patterns of moving objects while avoiding the obstacles in a picture. Then we classify them by DP matching against previously stored reference patterns in the database for all possible classes (person, bicycle, car, and bus).
In this paper, we compare several variations of an algorithm used to detect and classify objects passing laterally in front of a security camera. We test both a moving object speed independent and a speed based method for constructing patterns for the passing objects. The results show that the relatively simple method of pattern classification by DP matching can be successfully applied for classifying graphic objects of a certain degree of complexity. Finally, non-homogeneity of people's patterns and their subsequent frequent misclassification is addressed by not producing reference patterns for people, and differentiating them from correctly classified bicycles by the DP distance to the first candidate.

抄録全体を表示