• Journal of Broadcast Engineering
• /
• v.20 no.5
• /
• pp.728-737
• /
• 2015

Stitching algorithm obtain a descriptor of the feature points extracted from multiple images, and create a single image through the matching process between the each of the feature points. In this paper, a feature extraction and matching techniques for the creation of a high-speed panorama using video input is proposed. Features from Accelerated Segment Test(FAST) is used for the feature extraction at high speed. A new feature point matching process, different from the conventional method is proposed. In the matching process, by tracking region containing the feature point through the Mean shift vector required for matching is obtained. Obtained vector is used to match the extracted feature points. In order to remove the outlier, the RANdom Sample Consensus(RANSAC) method is used. By obtaining a homography transformation matrix of the two input images, a single panoramic image is generated. Through experimental results, we show that the proposed algorithm improve of speed panoramic image generation compared to than the existing method.

• Journal of Internet Computing and Services
• /
• v.15 no.6
• /
• pp.117-124
• /
• 2014

Cluttered background is a major obstacle in developing salient object detection and tracking system for mobile device captured natural scene video frames. In this paper we propose a context aware feature vector selection model to provide an efficient noise filtering by machine learning based classifiers. Since the context awareness for feature selection is achieved by searching nearest neighborhoods, known as NP hard problem, we apply a fast approximation method with complexity analysis in details. Separability enhancement in feature vector space by adding the context aware feature subsets is studied rigorously using principal component analysis (PCA). Overall performance enhancement is quantified by the statistical measures in terms of the various machine learning models including MLP, SVM, Naïve Bayesian, CART. Summary of computational costs and performance enhancement is also presented.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.5
• /
• pp.960-966
• /
• 2017

In this paper, we proposed an algorithm that can extract lesion by inputting a medical image. Feature points are extracted using SIFT algorithm to extract genetic training of medical image. To increase the intensity of the feature points, the input image and that raining image are matched using vector similarity and the lesion is extracted. The vector similarity match can quickly lead to lesions. Since the direction vector is generated from the local feature point pair, the direction itself only shows the local feature, but it has the advantage of comparing the similarity between the other vectors existing between the two images and expanding to the global feature. The experimental results show that the lesion matching error rate is 1.02% and the processing speed is improved by about 40% compared to the case of not using the feature point intensity information.

• Journal of Internet Computing and Services
• /
• v.13 no.4
• /
• pp.1-10
• /
• 2012

In this paper, an optimal feature selection method considering sematic of features, which is preprocess of document classification is proposed. The feature selection is very important part on classification, which is composed of removing redundant features and selecting essential features. LSA (Latent Semantic Analysis) for considering meaning of the features is adopted. However, a supervised LSA which is suitable method for classification problems is used because the basic LSA is not specialized for feature selection. We also apply GA (Genetic Algorithm) to the features, which are obtained from supervised LSA to select better feature subset. Finally, we project documents onto new selected feature subset and classify them using specific classifier, SVM (Support Vector Machine). It is expected to get high performance and efficiency of classification by selecting optimal feature subset using the proposed hybrid method of supervised LSA and GA. Its efficiency is proved through experiments using internet news classification with low features.

• Journal of IKEEE
• /
• v.25 no.2
• /
• pp.273-279
• /
• 2021

In this paper, we propose an algorithm that shortens the learning time by performing individual learning using partitioning the deep learning structure. The proposed algorithm consists of four processes: network classification origin setting process, feature vector extraction process, feature noise removal process, and class classification process. First, in the process of setting the network classification starting point, the division starting point of the network structure for effective feature vector extraction is set. Second, in the feature vector extraction process, feature vectors are extracted without additional learning using the weights previously learned. Third, in the feature noise removal process, the extracted feature vector is received and the output value of each class is learned to remove noise from the data. Fourth, in the class classification process, the noise-removed feature vector is input to the multi-layer perceptron structure, and the result is output and learned. To evaluate the performance of the proposed algorithm, we experimented with the Extended Yale B face database. As a result of the experiment, in the case of the time required for one-time learning, the proposed algorithm reduced 40.7% based on the existing algorithm. In addition, the number of learning up to the target recognition rate was shortened compared with the existing algorithm. Through the experimental results, it was confirmed that the one-time learning time and the total learning time were reduced and improved over the existing algorithm.

• IEIE Transactions on Smart Processing and Computing
• /
• v.4 no.4
• /
• pp.195-201
• /
• 2015

For median filtering (MF) detection in altered digital images, this paper presents a new feature vector that is formed from autoregressive (AR) coefficients via an AR model of the gradients between the neighboring row and column lines in an image. Subsequently, the defined 10-D feature vector is trained in a support vector machine (SVM) for MF detection among forged images. The MF classification is compared to the median filter residual (MFR) scheme that had the same 10-D feature vector. In the experiment, three kinds of test items are area under receiver operating characteristic (ROC) curve (AUC), classification ratio, and minimal average decision error. The performance is excellent for unaltered (ORI) or once-altered images, such as $3{\times}3$ average filtering (AVE3), QF=90 JPEG (JPG90), 90% down, and 110% up to scale (DN0.9 and Up1.1) images, versus $3{\times}3$ and $5{\times}5$ median filtering (MF3 and MF5, respectively) and MF3 and MF5 composite images (MF35). When the forged image was post-altered with AVE3, DN0.9, UP1.1 and JPG70 after MF3, MF5 and MF35, the performance of the proposed scheme is lower than the MFR scheme. In particular, the feature vector in this paper has a superior classification ratio compared to AVE3. However, in the measured performances with unaltered, once-altered and post-altered images versus MF3, MF5 and MF35, the resultant AUC by 'sensitivity' (TP: true positive rate) and '1-specificity' (FN: false negative rate) is achieved closer to 1. Thus, it is confirmed that the grade evaluation of the proposed scheme can be rated as 'Excellent (A)'.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.8
• /
• pp.397-402
• /
• 2007

In this paper, we apply a support vector machine(SVM) that incorporates an optimized nonlinear decision rule over different sets of feature vectors to improve the performance of statistical model-based voice activity detection(VAD). Conventional method performs VAD through setting up statistical models for each case of speech absence and presence assumption and comparing the geometric mean of the likelihood ratio (LR) for the individual frequency band extracted from input signal with the given threshold. We propose a novel VAD technique based on SVM by treating the LRs computed in each frequency bin as the elements of feature vector to minimize classification error probability instead of the conventional decision rule using geometric mean. As a result of experiments, the performance of SVM-based VAD using the proposed feature has shown better results compared with those of reported VADs in various noise environments.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.3
• /
• pp.71-81
• /
• 1995

In this paper, feature vector extraction methods and classification algorithms for the automatic classification of transient signals in underwater are discussed. A feature vector extraction method using wavelet transform, which shows good performance with small number of coefficients, is proposed and compared with the existing classical methods. For the automatic classification, artificial neural networks such as multilayer perceptron (MLP), radial basis function (RBF), and MLP-Class are utilized, where those neural networks as well as extracted feature vectors are combined to improve the performance and reliability of the proposed algorithm. It is confirmed by computer simulation with Traco's standard transient data set I and simulated data that the proposed feature vector extraction method and classification algorithm perform well, assuming that the energy of a given transient signal is sufficiently larger than that of a ambient noise, that there are the finite number of noise sources, and that there does not exist noise sources more than two simultaneously.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.1
• /
• pp.87-92
• /
• 1996

Transient signals in underwater show several characterisrics, that is, short duration, strong nonstationarity, various types of transient sources, which make it difficult to analyze and classify transient signals. In this paper, the feature vector extraction method for transient SOMAR signals is discussed by applying digital signal processing methods to the analysis of transient signals. A feature vector extraction methods using wavelet transform, which enable us to obtain better recognition rate than automatic classification using the classical method, are proposed. It is confirmed by simulation that the proposed method using wavelet transform performs better than the classical method even with smaller number of feature vectors. Especially, the feature vector extraction method using PR-QMF wavelet transform with the Daubechies coefficients is shown to perform well in noisy environment with easy implementation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.9
• /
• pp.1547-1553
• /
• 2008

This paper presents a method that generates various facial expressions of vector graphic character by using the simplified principle component vector. First, we analyze principle components to the nine facial expression(astonished, delighted, etc.) redefined based on Russell's internal emotion state. From this, we find principle component vector having the biggest effect on the character's facial feature and expression and generate the facial expression by using that. Also we create natural intermediate characters and expressions by interpolating weighting values to character's feature and expression. We can save memory space considerably, and create intermediate expressions with a small computation. Hence the performance of character generation system can be considerably improved in web, mobile service and game that real time control is required.