• IEMEK Journal of Embedded Systems and Applications
• /
• v.17 no.3
• /
• pp.129-138
• /
• 2022

Collecting and labeling sufficient training data, which is essential to deep learning-based visual inspection, is difficult for manufacturers to perform because it is very expensive. This paper presents a steel plate surface defect detection system with industrial-grade detection performance by training a small amount of steel plate surface images consisting of labeled and non-labeled data. To overcome the problem of lack of training data, we propose two data augmentation techniques: program-based augmentation, which generates defect images in a geometric way, and generative model-based augmentation, which learns the distribution of labeled data. We also propose a 4-step semi-supervised learning using pseudo labels and consistency training with fixed-size augmentation in order to utilize unlabeled data for training. The proposed technique obtained about 99% defect detection performance for four defect types by using 100 real images including labeled and unlabeled data.

• Journal of the Korean Geotechnical Society
• /
• v.21 no.5
• /
• pp.111-122
• /
• 2005

The data on the pile load tests performed on 35 large drilled shafts are analyzed to investigate the load-settlement characteristics of large drilled shafts embedded in bed rocks. Generally, the settlement of large drilled shafts embedded in bed rocks is too small to determine the ultimated load with application of the regulation in design code for either the total settlement or the residual settlement. Therefore, to determine the yield load of large drilled shafts embedded in bed rocks, p(load)-logS (settlement) curve method, which has been proposed originally for the driven pile, was applied to the investigation on the data of the pile load tests. This technique shows that the yield load can be determined accurately and easily rather than other conventional techniques such as P-S, logp-logS, S-logt, and P-S curve methods. An empirical equation is proposed to represent the relationship between pile load and settlement before the yield loading condition. And the settlement of piles was related with the depth embedded in rock as well as rock properties. Based on the investigation on the data of pile load tests, the resonable regulations f3r both the total settlement and the residual settlement are proposed to determine the yield load of large drilled shafts embedded in bed rocks.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.10 no.3
• /
• pp.179-188
• /
• 2015

We propose obstacle classification method based on 2D LIDAR(Light Detecting and Ranging) database. The existing obstacle classification method based on 2D LIDAR, has an advantage in terms of accuracy and shorter calculation time. However, it was difficult to classifier the type of obstacle and therefore accurate path planning was not possible. In order to overcome this problem, a method of classifying obstacle type based on width data of obstacle was proposed. However, width data was not sufficient to improve accuracy. In this paper, database was established by width, intensity, variance of range, variance of intensity data. The first classification was processed by the width data, and the second classification was processed by the intensity data, and the third classification was processed by the variance of range, intensity data. The classification was processed by comparing to database, and the result of obstacle classification was determined by finding the one with highest similarity values. An experiment using an actual autonomous vehicle under real environment shows that calculation time declined in comparison to 3D LIDAR and it was possible to classify obstacle using single 2D LIDAR.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.10 no.3
• /
• pp.119-127
• /
• 2015

In the IoT(Internet of Things) environment, methods that user can access sensor node directly to collect sensing data or manage sensor in a gateway have a limitations. To solve this problem, cloud based sensor network architectures are proposed. In this paper, we proposed CoAP based IoT architecture that a lightweight gateway is used for data gathering instead of using a heavy traditional one and users can request sensing data through IoT applications running in the cloud environment and analyze signaling message cost. By doing so, our system can reduce message cost compared to the traditional gateway based system.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.2 no.1
• /
• pp.24-30
• /
• 2007

According to development of wireless communication technologies, we need not only high data rate but low data rate system of low power consumption. This low data rate system is utilized in the field of home automation, health care, sensoring and monitoring, etc. IEEE 802.15.4 LR-WPAN system is the best choice for realizing ubiquitous networking system. In this paper SoC Architecture for IEEE 802.15.4 Low Rate WPAN is designed. IEEE 802.15.4 Low Rate WPAN system serves the functions and realization of home area network. We propose the SoC architecture for 868/915MHz frequency band of IEEE 802.15.4 Low Rate WPAN system. The key issue is to design SoC architecture which provides the function of Low Rate WPAN system to meet the requirement of IEEE 802.15.4 standards.

• Convergence Security Journal
• /
• v.9 no.2
• /
• pp.85-89
• /
• 2009

This study to solve these problems a search system has been designed by combining the two methods. Also the search and manage image data by handheld devices such as portable PDA or smart phone, a system has been designed application to an embedded system. Once this is used, multimedia data can be efficiently searched and utilized by handheld devices.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.769-770
• /
• 2008

Mani Memory DataBase(MMDB) system store their data in main physical memory and provide very high-speed access. Conventional database system are optimized for the particular characteristics of disk storage mechanism. Memory resident systems, on the other hand, use different optimizations to structure and organize data, as well as to make it reliable. This paper provides a brief overview on MMDBs and the results after evaluating the performance of our simple MMDB based on Embedded system.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.7 no.3
• /
• pp.121-127
• /
• 2012

An integrated crypto engine for encryption and decryption of AES algorithm based on unified data-path architecture is efficiently designed and implemented in this paper. In order to unify the design of encryption and decryption, internal steps in single round is adjusted so as to operate with columns after row operation is completed and efficient method for a buffer is developed to simplify the Shift Rows operation. Also, only one S-box is used for both key expansion and crypto operation and Key-Box saving expended key is introduced provide the key required in encryption and decryption. The functional simulation based on ModelSim simulator shows that 164 clocks are required to process the data of 128bits in the proposed engine. In addition, the proposed engine is implemented with 6,801 gates by using Xilinx Synthesizer. This demonstrate that 40% gates savings is achieved in the proposed engine, compared to individual designs of encryption and decryption engine.

• Proceedings of the IEEK Conference
• /
• 2003.07b
• /
• pp.959-962
• /
• 2003

At the processing capabilities and operating frequency of embedded system are growing, so is the needed data bandwidth to fully utilize the processing capability. The ability to transfer huge amount of data between the embedded core and external devices is required for efficient system operation. In this paper, the data communication architecture for the mixed-clock system is proposed. The dynamic priority adaptation algorithm for bus arbitration is proposed for bandwidth guarantee. The communication architecture that incorporates the proposed arbitration algorithm adapts the priority of communication components dynamically based on the information from FIFO. The experiments show that the measured bandwidth of each component traces the required bandwidth well compared to the other arbitration algorithms

• IEMEK Journal of Embedded Systems and Applications
• /
• v.4 no.2
• /
• pp.69-78
• /
• 2009

Localization is a very important technique for the mobile robot to navigate in outdoor environment. In this paper, the development of the sensor fusion algorithm for controlling mobile robots in outdoor environments is presented. The multi-sensorial dead-reckoning subsystem is established based on the optimal filtering by first fusing a heading angle reading data from a magnetic compass, a rate-gyro, and two encoders mounted on the robot wheels, thereby computing the dead-reckoned location. These data and the position data provided by a global sensing system are fused together by means of an extended Kalman filter. The proposed algorithm is proved by simulation studies of controlling a mobile robot controlled by a backstepping controller and a cascaded controller. Performances of each controller are compared.