• Journal of Biosystems Engineering
• /
• v.34 no.4
• /
• pp.278-285
• /
• 2009

Navigation system is applied in variety of fields including the simple location positioning, autopilot navigation of unmanned robot tractor, autonomous guidance systems for agricultural vehicles, construction of large field works that require high precision and map making process. Particularly utilization of GPS (Global Positioning System) is very common in the present navigation system. This study introduces a navigation system for autonomous field robot that travels to the pre-input path using GPS information. Performance of the GPS- based navigation is highly depended on its receiving rate because GPS receivers do not acquire any navigation information in the period between the refresh intervals. So this study presents an algorithm that improves an accuracy of the navigation by estimation the positional information during the blind period of a low rate GPS receiver. In fact the algorithm calculated the robot's heading in a 50 Hz rate, so the blind period of an 1 Hz GPS receiver is extensively covered. Consequently implementation of the algorithm to the GPS based navigation showed an improvement in guidance accuracy. The conventional field robot directly carried an expensive control computer and sensors onboard, therefore the miniaturization and weight reduction of the robot was limited. In this paper, the field robot carried only communication equipments such as GPS module, normal RC receiver, and bluetooth modem. This enabled the field robot to be built in an economic cost and miniature size.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.5
• /
• pp.275-280
• /
• 2013

In this paper, a new stereo object extraction algorithm using a block-based MSE (mean square error) algorithm and the configuration parameters of a stereo camera is proposed. That is, by applying the SSD algorithm between the initial reference image and the next stereo input image, location coordinates of a target object in the right and left images are acquired and then with these values, the pan/tilt system is controlled. And using the moving angle of this pan/tilt system and the configulation parameters of the stereo camera system, the mask window size of a target object is adaptively determined. The newly segmented target image is used as a reference image in the next stage and it is automatically updated in the course of target tracking basing on the same procedure. Meanwhile, a target object is under tracking through continuously controlling the convergence and FOV by using the sequentiall extracted location coordinates of a moving target.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.6
• /
• pp.754-760
• /
• 2012

This paper describes peak shaving algorithms as long-term cycle scheduling in the power management system (PMS) for MW-scale battery energy storage systems (BESS). The purpose of PMS is basically to manage the input and output power from battery modules placed in the systems. Assuming that an one-day ahead load curve is provided, off-line peak shaving algorithms can be employed, but applying the results of the off-line algorithm may result in the difference in the real-time performance because there is uncertainty in the provided load curve. This paper adopts fuzzy based LP (linear programming) algorithms for describing the peak shaving algorithm in PMS and discusses a solution technique and real-time operation strategies using the solution.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.11
• /
• pp.2150-2157
• /
• 2007

In this paper, we have proposed a self disease diagnosis system for ordinary persons to help the decision of access methods to a specialized medical management, and for medical specialities to discover new diseases and their symptoms easily, using verification of an individual#s health status by a series of processes performed by oneself. In the proposed self disease diagnosis system, illness is decided by 60 kinds of diseases selected using the report called #Diseases that Koreans take seriously# published by Ministry of Health & Welfare and medical contents called #Engel Pharm#, and also using 161 representative symptoms for the 60 kinds of diseases. An individual#s health information is extracted by diagnosis of one#s health status by a clustering of the 60 kinds of diseases using enhanced ART2 algorithm and input vectors from the results of questions for symptoms of each disease.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.9
• /
• pp.1682-1688
• /
• 2008

The Counterpropagation algorithm(CP) is a combination of Kohonen competition network as a hidden layer and the outstar structure of Grossberg as an output layer. CP has been used in many real applications for pattern matching, classification, data compression and statistical analysis since its learning speed is faster than other network models. However, due to the Kohonen layer's winner-takes-all strategy, it often causes instable learning and/or incorrect pattern classification when patterns are relatively diverse. Also, it is often criticized by the sensitivity of performance on the learning rate. In this paper, we propose an enhanced CP that has multiple Kohonen layers and dynamic controlling facility of learning rate using the frequency of winner neurons and the difference between input vector and the representative of winner neurons for stable learning and momentum learning for controlling weights of output links. A real world application experiment - pattern recognition from passport information - is designed for the performance evaluation of this enhanced CP and it shows that our proposed algorithm improves the conventional CP in learning and recognition performance.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.18 no.3
• /
• pp.155-162
• /
• 2008

In this paper we proposed a video watermarking algorithm based on a three dimension discrete wavelet transform (3D DWT) and direct spread spectrum (DSS). In the proposed method, the information watermark is embedded into followed frames, after sync watermark is embedded into the first frame. Input frames are divided into sub frames which are located odd row and even row. The sub frames are arranged as 3D frames, and transformed into 3D wavelet domain. In this domain the watermark is embedded using DSS. Existing video watermarking using 3D DWT is non-blind method but, proposed algorithm uses blind method. The experimental results show that the proposed algorithm is robust against frame cropping, noise addition, compression, etc. acquiring BER of 10% or below and sustains level of 40dB or above on the average.

• The Journal of the Korea Contents Association
• /
• v.22 no.10
• /
• pp.1-10
• /
• 2022

Visually impaired people have very difficulty using retail stores due to the absence of braille information on products and any other support system. In this paper, we propose a basic algorithm for a system that recognizes products in retail stores and explains them as a voice. First, the deep learning model detects hand objects and product objects in the input image. Then, it finds a product object that most overlapping hand object by comparing the coordinate information of each detected object. We determine that this is a product selected by the user, and the system read the nutritional information of the product as Text-To-Speech. As a result of the evaluation, we confirmed a high performance of the learning model. The proposed algorithm can be actively used to build a system that supports the use of retail stores for the visually impaired.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.10
• /
• pp.34-42
• /
• 2012

This paper analyses the energy saving basic power models in core IP networks, and proposes the adaptive buffer and burst scheme which is a possible energy saving method, and its implementation algorithm in core IP networks. Especially this paper describes the adaptive buffer and burst scheme dynamically varying the buffering interval B according to the input traffic volume of ingress router, and explains the operation principle of proposed scheme. This method is to adjust the buffering interval B according to input traffic volume of ingress router, that is increasing the interval B when input traffic volume is low, and decreasing the interval B when input traffic volume is high between some given interval regions. This method can gets the high energy saving effect as decreasing the transition number of idle/active in networks when input traffic volume is low, and decreasing the transition number of idle/active by the continuous of burst packets in transit router when input traffic volume is high. This paper shows the increasing of asleep rate for the energy saving of core IP networks and confirms the energy saving of core IP networks by the computer simulation. We confirmed that proposed method can be save the energy of IP networks by properly trade off network performances.

• Journal of KIISE:Information Networking
• /
• v.30 no.1
• /
• pp.134-142
• /
• 2003

We present an ATM switch architectures based on the multistage interconnection network(MIN) for the efficient multicast traffic control. Many of these applications require multicast connections as well as point-to-point connections. Muiticast connection in which the same message is delivered from a source to arbitrary number of destinations is fundamental in the areas such as teleconferencing, VOD(video on demand), distributed data processing, etc. In designing the multicast ATM switches to support those services, we should consider the fairness(impartiality) and priority control, in addition to the overflow problem, cell processing with large number of copies, and the blocking problem. In particular, the fairness problem which is to distribute the incoming cells to input ports smoothly is occurred when a cell with the large copy number enters upper input port. In this case, the upper input port sends before the lower input port because of the calculating method of running sum, and therefore cell arrived into lower input port Is delayed to next cycle to be sent and transmission delay time becomes longer. In this paper, we propose the cell splitting and group splitting algorithm, and also the fairness scheme on the basis of the nonblocking characteristics for issuing appropriate copy number depending on the number of Input cell in demand. We evaluate the performance of the proposed schemes in terms of the throughput, cell loss rate and cell delay.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.1
• /
• pp.124-132
• /
• 2002

In this paper, we present a vision algorithm and method for input image improvement and preprocessing of dented and raised characters on the sidewall of tires. we define optical condition between reflect coefficient and reflectance by the physical vector calculate. On the contrary this work will recognize the engraved characters using the computer vision technique. Tire input images have all most same grey levels between the characters and backgrounds. The reflectance is little from a tire surface. therefore, it's very difficult segment the characters from the background. Moreover, one side of the character string is raised and the other is dented. So, the captured images are varied with the angle of camera and illumination. For optimum Input images, the angle between camera and illumination was found out to be with in 90$^{\circ}$. In addition, We used complex filtering with low-pass and high-pass band filters to improve input images, for clear input images. Finally we define equation reflect coefficient and reflectance. By doing this, we obtained good images of tires for pattern recognition.