• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.4
• /
• pp.781-790
• /
• 2008

In order to improve on the divergence by noise convariance in the Kalman filter position estimation, we propose a method of position estimating through compensating the autonomous mobile robot's noise. Proposed method is the modified Kalman filter using neural network. It is prevented the divergence by the estimation of measurement noise covariance and system noise covariance. In order to verify the effectiveness of the proposed method, we performed simulations and experiments for position estimation. The results show that convergence and position error is reduced than the Kalman filter method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.252-255
• /
• 2014

Image treatment is becoming mainstream as the demand for image restoration has drastically increased in the digital era. But in the process of acquiring, transmitting and treating video data, the salt and pepper noise damages the image. One of the major methods used for restoring images are SMF(standard median filter), CWMF(center weighted median filter) and SWMF(switching weighted median filter), but these filters all leave a bit to be desired in terms of removing noise and preserving edge. Therefore, a transformed median filter is suggested through the algorithm presented for the restoration of damaged images.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.05a
• /
• pp.798-800
• /
• 2014

In recent years, according to the improvement of Digital image technology have been recently developed most of communication technology from multimedia communication service as well as image data transmission. But In the process of storing and transmitting noise is still generated in noise and the image degrades rapidly quality of a lot of image impulse noise. To eliminate this noise, SMF, CWMF, SWMF etc. The filters have been proposed to interfere with the noise characteristics of the filter are somewhat sufficient. Therefore, in this paper, in order to remove impulse noise is proposed a modified median filter. And impulse noise removal algorithms to confirm the existed PSNR(peak signal to noise ratio) from using conventional methods were compared.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.10a
• /
• pp.393-396
• /
• 2016

Digital image processing is the technical area of processing and analysis with intellectual and efficient ways, which has been commercialized in a variety of applications. However, the noise is occurred in the image data with multiple reasons and various studies have been performed to eliminate the noise. Generally, the types of noise vary by causes and forms, and composite noise is the representative one. Hence, the modified switching filter to process by types of noise was suggested to eliminate composite noise in the image effectively and to have excellent characteristics of edge conservation.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.260-263
• /
• 2006

최근 온라인 게임 산업이 커짐에 따라 이를 즐기는 유저도 급증하고 있다. 온라인 게임에서는 일반적으로 유저들이 서로를 구분하기 위해 사용하는 사용자 이름과 상호간 의사소통을 하기 위한 채팅을 지원한다. 유저의 수가 증가함에 따라 대화의 양은 더욱 더 많아지고, 선정성, 폭력성을 띄는 언어의 문제로 이어지고 있다. 이는 특히 18세 이하도 이용가능한 게임을 만드는 경우 더욱 중요하다. 하지만 대부분의 게임들이 금지어 리스트에 따른 단어 매칭방식의 비속어 필터링만을 제공하고 있다. 이러한 방법은 금지어로 지정된 단어를 포함한 정상적인 채팅도 막을 뿐만 아니라 일부 음절을 다른 기호로 바꾸어 표기한 비속어는 걸러내지 못한다. 변형된 단어들을 충분히 처리하지 못한다면 비속어 필터링 시스템은 단지 무력하고 쓸모없는 존재가 될 뿐이다. 본 논문에서는 SVM을 이용하여 학습이 가능한 비속어 필터링 시스템을 제안하고자 한다. SVM을 이용하면 사용자 편의성을 해치지 않고서도 보다 많은 종류의 비속어들을 효과적으로 걸러낼 수 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04a
• /
• pp.748-750
• /
• 2002

본 논문에서는 이미지 필터 모듈을 CORBA를 사용하여 분산 엔진으로 구현하는 방법에 대한 연구를 수행하였다. 이미지 필터 모듈들은 영상 처리 과정에서 기본적으로 응용되는 전처리 모듈로서 원 영상을 목적에 따라 변형시키는 기능을 수행하며, 다양한 어플리케이션에서 여러 플렛폼으로 빈번하게 구현되곤 한다. 이 연구에서는 각각 독립 적으로 구현된 이미 지 필터 모듈을 CORBA를 사용한 분산 서버로 구축함으로서 , 다양한 종류의 플랫폼에서 구현된 여러 어플리케이션들이 이미지 필터 기본 모듈들을 별도의 구현 없이 쉽게 사용할 수 있도록 하였다.00

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.05a
• /
• pp.397-400
• /
• 2016

In digital images, the addition due to noise occurs in the process of obtaining, saving, and transmitting. For examples of noise, there are salt and pepper noise, Gaussian noise, and composition noise where various noises are mixed. Existing filters have insufficient noise removal characteristics because it uses single filters in composite noise environment. Therefore the study suggested a switching filter that processes with special weighted value and median filter according to local mask salt and pepper noise density when central pixel is damaged by salt and pepper noise, and processes by applying weighted values differently according to standard deviation of local mask when damaged by Gaussian noise.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.05a
• /
• pp.792-794
• /
• 2014

In the modern society where the communication technology has rapidly developed, image devices such as digital display, camera, etc., forms the center. However, during the transmission of image data, storing, and obtaining, a noise is added to the image due to various reasons and degrades the quality of the image. In this paper, an average filter algorithm modified in order to ease the effect of AWGN(additive white Gaussian noise) being added to the image was proposed. Also compare existing methods through the using PSNR.

• The KIPS Transactions:PartB
• /
• v.15B no.1
• /
• pp.61-68
• /
• 2008

This paper presents a spam filter system using co-training with spamminess features and URL features based on the maximum entropy model. Spamminess features are the emphasizing patterns or abnormal patterns in spam messages used by spammers to express their intention and to avoid being filtered by the spam filter system. Since spammers use URLs to give the details and make a change to the URL format not to be filtered by the black list, normal and abnormal URLs can be key features to detect the spam messages. Co-training with spamminess features and URL features uses two different features which are independent each other in training. The filter system can learn information from them independently. Experiment results on TREC spam test collection shows that the proposed approach achieves 9.1% improvement and 6.9% improvement in accuracy compared to the base system and bogo filter system, respectively. The result analysis shows that the proposed spamminess features and URL features are helpful. And an experiment result of the co-training shows that two feature sets are useful since the number of training documents are reduced while the accuracy is closed to the batch learning.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.7
• /
• pp.1373-1379
• /
• 2007

A modified adaptive switching median filter for impulse noise removal, which has the noise detection step and the noise filtering step, is proposed in this paper. In the noise detection step, we use the detection threshold which is earned by calculating the intensity differences between pixels nearby with each other in localized window, to determine whether the pixels in the image are noise or not. Then in the noise filtering step, we will only remove the corrupted pixels and remain the good pixels. By the noise detection result, we can easily get the local noise density of the image, and use it to consider the filtering mask size and the times of filtering iteration according to different localized noise corruptions. For Setting the simulation result, we compared the proposed method to conventional median filters with several test images corrupted by various impulse noise densities. We also use the peak signal-to-noise ratio (PSNR) to evaluate restoration performance, the simulation results demonstrate that the proposed method shows better results than other median-based type filters.