Hosei University Repository
Not a member yet
22008 research outputs found
Sort by
DEVELOPMENT OF A RAPID PREDICTION METHOD OF CERAMIC GREEN DENSITY FOR SHEET CASTING PROCESS
Get PDFWe reported that the gravitational settling test can be suitable for prediction of the green density for sheet casting process because the settling behavior represents the thickening behavior of slurry. [1] However, the problem is that it takes a long time to perform a sedimentation test in gravitational field. Therefore, in this study, we investigated whether the density of the green sheet could be predicted by two methods: ①centrifugal sedimentation test and ②drying test in which the slurry was dried in a cup and the packing fraction was calculated from the initial slurry height and final powder bed height. It was shown that the packing fraction of the dried sediment obtained from the drying test had a good connection to the green sheet density, while the packing fraction of the sediment obtained from the centrifugal sedimentation test didn’t. It was demonstrated that the characterization time of slurry packing ability in order to predict the green sheet density could be shorten by using the drying test
IMITATION OF MAMMALIAN COCHLEA BASED ON ASYNCHRONOUS CELLULAR AUTOMATON
Get PDFIn this paper, we analyze the frequency response characteristics and tuning curve of an asynchronous cellular automaton cochlear model, and search for parameters that can reproduce the frequency characteristics of the cochlea of living and dead guinea pigs and the tuning curve of a bat. In addition, the model is implemented in FPGA, and the hardware cost is compared to an ODE cochlear model
MODEL-BASED OPTIMAL VIBRATION STIMULUS FOR KINESTHESIA PRESENTATION BASED ON THE RELATIONSHIP BETWEEN JOINT ANGLE AND MUSCLE ELONGATION
Get PDFMost of the existing studies on kinesthesia presentation by vibratory stimulation of tendons use a frequency corresponding to the angular velocity of the joint. However, the primary terminal nerves that fire in response to vibration stimulation are essentially those that fire in response to muscle elongation and growth rate. In this study, we performed the same task with the vibration frequency corresponding to the joint angular velocity and the muscle length change, and investigated the optimal vibration stimulus through a comparison of kinesthesia generation
JAPANESE COINS AND BANKNOTES RECOGNITION FOR VISUALLY IMPAIRED PEOPLE
Get PDFRecent deep learning techniques are successfully integrated into devices to assist visually impaired people in their daily lives, particularly detecting coins/banknotes. Previous works have focused on well captured devices and examined high-quality images. In this work, we design a framework to recognize Japanese Coin/Banknote (JCB) for low-quality images under various criteria. Discriminate features usually disappear in low-quality images. Consequently, using the depth image in addition to RGB image in processing can be enhanced the accuracy of our system. In this work, we first leverage depth information by using a Monocular Depth Prediction network. Additionally, a pre-trained Deep Convolutional Neural Network process RGB and Depth images, respectively. At last, we combine two networks by an ensemble method to produce more accurate detections. By processing depth images in addition to RGB images, the detection results are thus accurate. As a result, our work achieves 74.1% mean Average Precision (mAP)
Bi-directional Intra Prediction Based Measurement Coding and Encoder Rate Control for Compressive Sensing Images
Get PDFThis work presents a bi-directional intra prediction-based measurement coding algorithm to reduce spatial redundancies found in block-based compressive sensing schemes. By exploiting the fixed basis of a structured measurement matrix that allows members in measurement to be traceable to the position of the pixel, we choose the closest element between neighboring to make the prediction candidates. Following that, we apply the triangle method to scalar quantization, resulting in a controllable encoding bit rate that maximizes performance per block against bit rate consumption. Experiment results show that this work improves PSNR by 0.01 - 0.02 dB and bit rate reductions by 19% on average and up to 36% compared to the state-of-the-art
RELATIONSHIP BETWEEN DIURNAL FLUCTUATION RHYTHM OF HEART RATE AND POSITIVE PSYCHOLOGICAL INDEX
Get PDFThis paper proposes a method to associate IoT based physiological indices with psychological flow and related mental states defined in the positive psychology. The following physiological indices were used for the analysis. They are the double cosinor heart rate rhythm parameters, i.e., 24-hour and 12-hour period component amplitude A_24 and A_12, combined heart rate rhythm amplitude Camp, autonomic switching rate newly defined original parameter ASR. Psychological indices are obtained from questionnaires to evaluate subjects’ psychological dispositional characteristics. Three groups of subjects consist of university students US (N=33, aged 19-22), system engineers SE (N=20, aged 25-57), and senior people SP (N=30, aged 42-91). Regression analysis of the Flow Frequency showed the multiple regression of (A_24, ASR) and (Camp, ASR, A_12) gave the highest adjusted R^2 values for SE (R^2=0.366, p=0.052) and US (R^2=0.706, p=0.005) respectively, while the single regression of A_12 gave the highest adjusted R^2 value for SP (R^2=0.235, p=0.110). The research revealed the positive psychological tendency has been associated with the diurnal heart rate rhythm parameters. The method developed in this study may open the new direction in the ubiquitous mental health monitoring with the advances of IoT wrist device for the continuous heart rate record
ACCELERATION OF A PHOTON TRANSPORTATION WITH A GPU AND ITS APPLICATION TO MAMMOGRAPHY IMAGES
Get PDFLow-energy X-rays used in mammography cause a large effect by coherent scattering, and the resulting scattered photons decrease the quality of mammograms in terms of the contrast resolution. When we calculate these effects by using a Monte Carlo simulation with a general purpose code such as EGS, the complexity of the phenomenon increases the simulation time. To solve this problem, we developed a parallel processing code using a GPU and compared the performance of our code with these of EGS5. As a result, we succeeded in achieving 23 times faster computation with the same accuracy as EGS5. Then we applied our code to calculate the scattered photons with a numerical simulation phantom, and tried to estimate the scattered photons with a machine learning method. The results showed that we can estimated the scattered photons with a MSE of about 3.3%
RESEARCH ON ELECTRO-OPTICAL PROBES USING KTN CRYSTALS
Get PDFThis paper describes the feasibility of an electro-optic (EO) probe system using KTa_1-xNb_xO_3 (KTN) crystals. The application of EO probe system can be expand by using KTN crystal that responds to low frequency signals including direct current. We investigated the time stability of the measured values required for the sensor. We found that the tendency for time stability depends on the presence or absence of external light to the KTN. In addition, we raised the issue of drift, where the measurements of the two photodiodes change in the same direction
CONSIDERATION Of BOUNDARY CONDITIONS OF LATTICE BOLTZMANN METHOD
Get PDFWhen the fluid has a certain velocity, it becomes a turbulent flow in which a vortex is generated from the laminar flow. In turbulence, it is necessary to have a model corresponding to each case due to a vortex consisting of large and small, and a boundary condition that reduces the error from the theoretical value derived from the Navier-Stokes equation. Further, when a turbulent flow occurs, the calculation load increases near the boundary, so that the calculation time is required to be shortened. Therefore, the calculation time is shortened by performing parallel processingin a plurality of memories
Development of Virtual 2D-LIDAR Module by using 3D-LIDAR Point Cloud
Get PDFNowadays， 3D-LIDAR is one of the key essential devices to understand the profile of the surrounding environment for safe navigation of autonomous mobile robots． The 3D-LIDAR can capture a 360-degree surrounding profile at once as Point Clouds． However， when a 3D-LIDAR is attached to the mobile robot， many 3D-LIDARs have a blind spot especially near surrounding the mobile robot area． Therefore， most developed mobile robots use both 3D-LIDAR and 2D-LIDAR to reduce the blind spot area and prevent collisions with bumps and/or dynamic moving obstacles． In this paper， we describe a new Virtual 2D-LIDAR Module by using 3D-LIDAR Point Cloud data without actually implementing 2D-LIDAR． Once， the mobile robot capture near the surrounding profile by using 3D-LIDAR， 2D-LIDAR data can be generated which can set a virtual place on the mobile robot． Once the mobile robot acquires the surrounding environment profile using 3D-LIDAR， 2D-LIDAR can be set up virtually at any location on the mobile robot and the data can be acquired． The effectiveness of the developed Virtual 2D-LIDAR Module has been implemented as a ROS module and has been confirmed in actual preliminary experiments