Makishi et al. (2017) proposed a method that estimates the parameters for a bilateral filter, based on the distribution distance of images. This study proposes an improved version of the Makishi et al. (2017) method, which includes three new contributions. Firstly, our improved method was verified with two image sizes and two noise distributions. Secondly, five distribution distance functions were examined for the local search of parameter sets, where SKLD was found to be slightly superior to HD in terms of image quality and execution time. Finally, the parallelization of local search was examined for faster parameter estimation with recent multi-core processors. With a 2.8GHz Xeon processor of 10 cores, the parameter estimation times were reduced to 36ms (256x256 pixel) and 1.03 s (1024x1024 pixel).
Obfuscation is a method to conceal the structure and function of software. Obfuscator-LLVM (oLLVM) is a set of tools for software obfuscation, which is implemented as middle-end passes in an LLVM framework. This study investigates the hardware obfuscation scheme, where the C-codes obfuscated by oLLVM are converted into logic design by using an HLS (High Level Synthesis) tool. CHStone benchmark suite is used for evaluation, where each application is processed by oLLVM, C-backend, and Xilinx Vivado HLS to generate the corresponding obfuscated logic design. On average, the logic scales of obfuscated designs became 1.1 times larger (by bogus control-flow), 1.7 times larger (by control-flow flattening), and 1.2 times larger (by instruction substitutions). In some applications, the obfuscations added by oLLVM were canceled by the analysis and optimization of HLS.
The satisfiability (SAT) problem is widely applicable to various industrial problems, such as formal verification of industrial systems. This study presents the performance evaluation of a parallel SAT solver (Glucose Syrup) on Intel Xeon Phi many-core processors. After limiting the number of learnt clauses to be shared, the number of solved problems in the SAT Competition 2016 workload increased from 201 to 236, yet it was smaller than that of a Xeon processor.
We developed a mobile learning system for fishery and ocean techniques using video contents. The system aims to improve the current situation in the field of education on fishery and ocean for skill succession. The proposal is based on the industry-school education process for skill succession. By accessing the web site and watching practical videos, the users can learn methodologies on fishery and ocean. As a result, we showed that the system is effective for sharing and succession of techniques for the learner. Application of the proposed system in the class showed improvement in the students' autonomous learning.
In this study, we developed a current sensorless voltage driven magnetic levitation system with a Hall element displacement sensor using Kalman filter. The system costs less and has higher portability than the current driven system developed in a previous study, and is utilized effectively for model-based development education. However, the use of the displacement sensor for a voltage driven magnetic levitation system causes two problems, namely compensation for measurement error and correction of Hall element output voltages in the sensor. Therefore, we have proposed compensation and correction methods using Kalman filter. To verify the validity of the proposed methods, a magnetic levitation control experiment with the developed system was conducted.
In this paper, a design technique for ternary logic function circuits based on CMOS design are proposed. The circuits operate as B-ternary (binaric ternary) logic functions that are useful in asynchronous systems or self-checking circuits of fault-tolerant systems. Further, a special logic circuit, D-element, is proposed. D-element is designed for asynchronous systems like Mullers C-element in binary circuits. The circuits' SPICE simulations are provided to show their efficacy.
An alternative learning and its application to construct an overviewing human detection system (OHDES-V2) of flying drone for emergency rescue and investigation is presented in this paper. In this system, a deep neural network and alternative learning are used key techniques for object recognition from a free viewpoint. Simple appearance-based characteristics is determined from captured images, and the system uses a deep neural network to automatically classify human body, automobiles and so forth. The proposed system shows that several objects can be recognized from a bird's-eye view. Experimental results show that the system can effectively recognize four types of objects and walking persons with accuraces of 98.5% and 97.12%, respectively.
This paper proposes a new lazy learning algorithm, named balanced-kNN, for high performance robust classification of noisy patterns. K-nearest neighbor (k-NN) is a simple and powerful method with a high accuracy for various real world applications using unbiased datasets. However, noisy datasets are often gathered in real world applications. This paper presents a new robust algorithm, balanced-kNN, and compares the prediction accuracy with some conventional methods by using UCI datasets. The experimental results show that the balanced-kNN algorithm can perform more efficient classification of noisy data than the normal-kNN and weighted-kNN algorithms.
Mango is a popular seasonal fruit in summer, and mango production in the Okinawa prefecture is the largest in Japan. However, the growth process of mangoes is susceptible to changes in sunshine, temperature, and other factors. In this research, to produce mangoes without being influenced by the natural environment, we will develop a production system that can positively realize environmental measurement and control by using an IoT sensing system. Specifically, we introduced local CO2 application technology and a supplemental LED lighting system with the aim of activating photosynthesis by mango trees.
Crop growth is influenced by various environmental factors. As a condition for effective plant growth, it is necessary that control of the habitat environment, such as soil moisture content, temperature, humidity, solar radiation amount, is adequately delivered. We focused the fact that soil moisture is important. It is desirable that moisture is uniformly distributed in the soil. In order to avoid physiological disorders caused by water related stress on agricultural crops, it is necessary to use soil with appropriate permeability. By calculating and estimating some characteristic amount related to water permeability, it is possible to identify soil with appropriate water permeability. In this research, we realize the visualization of the planar soil moisture volume by arranging the soil moisturemeter at multiple points. In addition to a system that realizes planar visualization of soil moisture, we propose a method for estimating soil permeability parameters in combination with computer simulation.
The mainstream of conventional underwater radio communication technology is acoustic communication; however, there is a problem in that the communication speed is limited. On the other hand, optical wireless communication technology has been utilized, wherein in which large capacity data transmission becomes possible. In this paper, we show the configuration of the transmitter and receiver for FPGA mounted video and its evaluation results for video transmission by underwater visible light communication. As a result of the evaluation, it is shown that a transmission speed of 20Mbps or more can be achieved, and the total power consumption of the visible light communication transceiver was at most 0.372W.
In this paper, we propose a reconfigurable fault tolerant architecture that can recover from failure status with spare space. Recently, progress in semiconductor technology has been remarkable due to microfabrication of devices. The semiconductor technique plays an important role in artificial satellites and aircraft. Furthermore, it has guaranteed the reliability of the circuits by the multiplexing structure. However, in embedded systems, space-saving is regarded to be as important as reliability. In the traditional approach, the area overhead tends to become large. Reconfigurable fault tolerance can achieve high area efficiency. In this research, we aim to improve the reliability and area efficiency for a single stuck-at fault of the processor. In this article, we reproduce the proposed method using Tcl script and proposed standalone fault tolerant operation using embedded Linux.
Coral plays an important role in the marine ecosystem. They provide spawning sites and living spaces for fish, and control the concentration of the oxygen and carbon dioxide. However nowadays the percentage of living coral is getting less than the past, due to human-related activities. There are several researches on conservation, restoration, displacement, etc., of coral leaf. In order to conduct these researches appropriately, coral eggs are essential. However, it needs a lot of time and cost to obtain coral eggs, and there is a lot of burden on researchers. To solve this problem, we propose an un-manned full automatic air boat robot using a drone flight controller (FC) to collect coral eggs from the sea. This paper describes the prototype implementation of our air boat.
The authors are developing a condition monitoring system using vibration analysis and machine learning for the purpose of monitoring the condition of railway vehicle equipment. In railway vehicles, vibrations change due to long-term state change, so long-term data should be used for learning. In this case, it is not practical to use all data, so it is necessary to use only some part of the data, which is called prototype data. Therefore, a prototype selection method based on the neighborhood method is proposed in this paper. As a result of applying the proposed method to the vibration data during the abnormal simulation test, the expected effect was confirmed.
One of the problems in urban railway lines in the Tokyo area is that minor delays occur quite often during morning rush hours. To prevent these minor delays from occurring, railway companies are attempting to employ various types of countermeasures. However, these countermeasures may be extremely expensive to implement and thus, it is critical to quantitatively evaluate the effectiveness of these countermeasures. To this end, train traffic simulations are used. In this paper, we introduce a novel simulation approach that uses historical train traffic records. In addition, we introduce a concept to determine the arrival times of trains by using primarily the running times whereas in existing research the arrival times were determined from the headway time. We determine a rule to estimate the running times from the interval of trains at the previous station. We validated our simulator by comparing the simulation results with actual train traffic data.
This paper describes the torque and power factor characteristics of SynRMs (synchronous reluctance motors) considering their magnetic energy and co-energy properties. Since the inductance of SynRM varies drastically due to the magnetic saturation effect, it is not easy for us to understand their characteristics, MTPA (maximum torque per ampere) points, and maximum power factor points, quantitatively applying the frame based on equivalent circuit models. In this study, we focused on the fact that the model constructed by the magnetic energy and co-energy gives more simple expressions to the torque and power factor compared to the equivalent circuit model. A SynRM that consists of a stator, same as a benchmark model (D-model) provided by IEEJ, and a rotor with 3 layer flux-barrier was simulated by the electromagnetic field analysis, and its magnetic energy and co-energy were calculated. As a result, three states (I, II and III) were defined depending on degrees of the magnetic saturation level. Then, the output torque per square ampere becomes maximum on a boundary region between state II and III. Further, the current phase of the MTPA point moves to the leading phase side with increasing the current magnitude on state III and the phase of the maximum power factor point is larger than that of the MTPA point. It was clarified that the model constructed by the magnetic energy and co-energy are suitable for discussions on the SynRM characteristics taking account of the magnetic saturation.