• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.6
• /
• pp.604-608
• /
• 2008

The TFT-LCD image allows non-uniform illumination variation and that is one of main difficulties of finding defect region. The SQI (self quotient image) has the HPF (high pass filter) shape and is used to reduce low frequency-lightness component. In this paper, we proposed the TFT-LCD defect-enhancement algorithm using characteristics of the SQI, that is the SQI has low-frequency flattening effect and maintains local variation. The proposed method has superior flattening effect and defect-enhancement effect compared with previous the TFT-LCD image preprocessing.

• Journal of Digital Contents Society
• /
• v.10 no.1
• /
• pp.121-127
• /
• 2009

The automotive ECU(Electronic Control Unit) becomes more complicated and is demanding many functions. For example, many automobile companies are developing driver convenience systems such as power window switch, LCM(Light Control Module), mirror control system, seat memory. In addition, many researches and developments for DIS(Driver Information System) are in progress. It is dangerous to operate such systems in driving. In this paper, we implement the speech recognition system which controls the car convenience system using speech, and apply the preprocessing filter to improve the speech recognition rate in car noise environment. As a result, we get the good speech recognition rate in car noise environment.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38B no.7
• /
• pp.581-588
• /
• 2013

In the basic GMSK, the modulated signals are generated by the preprocessing of NRZ pulse trains with the gaussian low pass filter. In this paper, we proposed the GMSK modulation system with the delta pulse trains. The proposed system constructed with the gaussian pulse generator triggered by the bipolar delta pulse, the differentiator between $+T_b/2$ and $-T_b/2$, and then a matched filter for frequency-modulation function. The spectrum of proposed system was found to be smaller than about 35[%] at 40[dB] attenuated bandwidth, and the center spectrum to be lower than the basic GMSK with zero-crossing property.

• Korean Journal of Remote Sensing
• /
• v.14 no.2
• /
• pp.137-148
• /
• 1998

An algorithm developed for the precision correction of high resolution satellite images is introduced in this paper. In general, the polynomial warping algorithm which derives polynomial equations between GCPs extracted from an image and a base map requires many GCPs well-distributed over the image. The precision correction algorithm described in this paper is based on a sensor-orbit-Earth geometry, and therefore, it is capable of correcting a raw image using only 2-3 GCPs. This algorithm estimates the errors on the orbit determination and the attitude of the satellite by using a Kalman filter. This algorithm was implemented, tested and integrated into the KITSAT-3 image preprocessing software.

• Smart Media Journal
• /
• v.7 no.4
• /
• pp.24-29
• /
• 2018

Text analysis technique for natural language processing in deep learning represents words in vector form through word embedding. In this paper, we propose a method of constructing a document vector and classifying it into spam and normal text message, using word embedding and deep learning method. Automatic spacing applied in the preprocessing process ensures that words with similar context are adjacently represented in vector space. Additionally, the intentional word formation errors with non-alphabetic or extraordinary characters are designed to avoid being blocked by spam message filter. Two embedding algorithms, CBOW and skip grams, are used to produce the sentence vector and the performance and the accuracy of deep learning based spam filter model are measured by comparing to those of SVM Light.

• Journal of Information Processing Systems
• /
• v.2 no.1
• /
• pp.13-17
• /
• 2006

In this paper, we propose an audio extraction method to decrease the influence of the original signal by modifying the watermarking detection system proposed by P. Bassia et al. In the extraction of the watermark, we employ a simple mean filter to remove the influence of the original signal as a preprocessing of extraction and the repetitive insertion of the watermark. As the result of the experiment, for which we used about 20 kinds of actual audio data, we obtain a watermark detection rate of about 95% and a good performance even after the various signal processing attacks.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2007.04a
• /
• pp.315-319
• /
• 2007

This paper presents a algorithm, what is able to recognize 4 different continuous human motion using a single stationary camera as input. For the first step, we acquire images from a camera. To enhance the image, we perform preprocessing which deals with removing noise using median filter, thresholding. And then morphological operations are performed to remove which small blobs and eliminates small holes. At the forth step, blobs are analysed to extracts for foreground region. Then, motions are predicted from these images by using optical flow technique, and the predicted motion data are refined by comparing our cardboard models so as to judge behavior pattern.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.4
• /
• pp.1815-1821
• /
• 2015

Rapid increase in population growth has made the mankind to delve in appropriate identification of individuals through biometrics. Foot Print Recognition System is a new challenging area involved in the Personal recognition that is easy to capture and distinctive. Foot Print has its own dimensions, different in many ways and can be distinguished from one another. The main objective is to provide a novel efficient automated system Segmentation using Foot Print based on structural relations among the features in order to overcome the existing manual method. This system comprises of various statistical computations of various foot print parameters for identifying the factors like Instep-Foot Index, Ball-Foot Index, Heel- Index, Toe- Index etc. The input is naked footprint and the output result to an efficient segmentation system thereby leading to time complexity.

• Journal of information and communication convergence engineering
• /
• v.7 no.4
• /
• pp.535-538
• /
• 2009

Speckle noise reduction for ultrasound CT image using morphological adaptive median filtering based on edge preservation is presented in this paper. Speckle noise is multiplicative feature and causes ultrasound image to degrade widely from transducer. An input image is classified into edge region and homogeneous region in preprocessing. The speckle is reduced by morphological operation on the 2D gray scale by using convolution and correlation, and edges are preserved. The adaptive median is processed to reduce an impulse noise to preserve edges. As the result, MAM of the proposed method enhances the image to about 10% in comparison with Winner filter by Edge Preservation Index and PSNR, and 10% to only adaptive median filtering.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.1023-1024
• /
• 2008

This paper proposes a pedestrian counting system using line laser. By using a line laser and IR filter, the shapes of pedestrians are easily obtained without complex preprocessing. Also, the directions of pedestrians were able to distinguish by employing gradient information. In the experiment, the proposed method successfully counted the number of people with accuracy of about 97% and with processing time of 24ms per frame.