• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.3
• /
• pp.1319-1324
• /
• 2012

Tomatoes were harvested at different ripening stages. To determine the ripening stages, We analyzed the relation between the compound concentrations of tomato measured with HPLC and the tomato RGB images. Among the compound concentrations, tomato quality is mostly affected by the Lycopene. The $Q^2$ error of the predicted Lycopene concentration and the corresponding independent component of tomato RGB image, determined from the PLS procedure, was 0.92. and we show the effectiveness of the independent component by comparing the error between the pixel area of RGB image applied by independent component and the simple black white tomato image. This regression made it possible to construct concentration images of the tomatoes, which showed non-uniform ripening. The method can be applied in an unsupervised real time sorting machine of unripe and discolored tomato using the compound concentrations.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.05a
• /
• pp.401-404
• /
• 2006

본 논문에서는 영상의 중심이동과 독립성분분석에 의한 효율적인 표정 인식방법을 제안하였다. 여기서 중심이동은 얼굴영상의 1차 모멘트에 의한 전처리 과정으로 불필요한 배경을 배제시켜 계산시간의 감소 및 인식률을 개선하기 위함이다. 또한 독립성분분석은 얼굴표정의 특징으로 기저영상을 추출하는 것으로 고차의 통계성을 고려한 중복신호의 제거로 인식성능을 개선하기 위함이다. 제안된 방법을 320*243 픽셀의 48개(4명*6장*2그룹) 표정을 대상으로 Euclidean 분류척도를 이용하여 실험한 결과, 전처리를 수행치 않는 기존방법에 비해 우수한 인식성능이 있음을 확인하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.11b
• /
• pp.627-630
• /
• 2003

본 연구에서는 고정점 알고리즘의 독립성분분석과 원 신호의 시간적 상관성을 고려한 견실 알고리즘의 독립성분분석을 각각 이용하여 혼합지문영상을 분리하였다. 여기서 고정점 알고리즘은 뉴우턴법의 경신규칙을 이용함으로써 빠른 분리속도를 가진다. 견실 알고리즘은 2차적 통계성의 일괄처리 알고리즘으로 시간적 상관성과 낮은 kurtosis를 가진 영상분리에 효과적이다. 이들 기법들을 $256{\times}256$ 픽셀의 8개 지문으로부터 임의의 혼합행렬에 따라 발생되는 지문의 분리에 적용한 결과, 견실 알고리즘이 고정점 알고리즘의 독립성분분석에 비해 우수한 분리성능과 빠른 분리속도가 있음을 확인하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.11b
• /
• pp.623-626
• /
• 2003

본 연구에서는 고정점 알고리즘의 독립성분분석과 원신호의 시간적 상관성을 고려한 견실 알고리즘의 독립성분분석을 혼합한 조합형 독립성분분석에 의한 혼합영상의 분리를 제안하였다. 여기서 고정점 알고리즘은 뉴우턴법의 경신규칙을 이용함으로써 빠른 분리속도와 우수한 분리성능을 가지며, 견실 알고리즘은 2차적 통계성의 일괄처리 알고리즘으로 시간적 상관성 및 낮은 kurtosis를 가진 영상분리에 효과적이다. 이들 기법들을 $512{\times}512$ 픽셀의 4개 영상으로부터 임의의 혼합행렬에 따라 발생되는 흔합영상의 분리에 적용한 결과, 우수한 분리성능과 빠른 분리속도가 있음을 확인하였다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.90-93
• /
• 2004

본 논문에서는 고정점 알고리즘의 독립성분분석을 이용한 물체영상의 특징추출을 제안하였다. 여기서 고정점 알고리즘은 뉴우턴법에 기초한 것으로 빠른 특징추출성능을 얻기 위함이고, 독립성분분석의 이용은 통계적으로 독립인 기저영상을 효과적으로 추출하기 위함이다. 제안된 기법을 Image＊after사에서 제공하는 352$\times$264 픽셀의 10개 물체영상을 대상으로 실험한 결과, 빠르면서도 정확한 복원성능과 PCA보다도 개선된 특징 추출성능이 있음을 확인하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.7
• /
• pp.539-544
• /
• 2003

This paper describes how to reduce the effect of an occupation threshold by that the transform of mixture components of HMM parameters is controlled in hierarchical tree structure to prevent from over-adaptation. To reduce correlations between data elements and to remove elements with less variance, we employ PCA (Principal component analysis) and ICA (independent component analysis) that would give as good a representation as possible, and decline the effect of over-adaptation. When we set lower occupation threshold and increase the number of transformation function, ordinary MLLR adaptation algorithm represents lower recognition rate than SI models, whereas the proposed MLLR adaptation algorithm represents the improvement of over 2% for the word recognition rate as compared to performance of SI models.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.10
• /
• pp.48-58
• /
• 1998

We present a method that can recognize face images using independent component analysis that is used mainly for blind sources separation in signal processing. We assumed that a face image can be expressed as the sum of a set of statistically independent feature images, which was obtained by using independent component analysis. Face recognition was peformed by projecting the input image to the feature image space and then by comparing its projection components with those of stored reference images. We carried out face recognition experiments with a database that consists of various varied face images (total 400 varied facial images collected from 10 per person) and compared the performance of our method with that of the eigenface method based on principal component analysis. The presented method gave better results of recognition rate than the eigenface method did, and showed robustness to the random noise added in the input facial images.

• The KIPS Transactions:PartB
• /
• v.13B no.1 s.104
• /
• pp.7-14
• /
• 2006

This paper presents an efficient face recognition method using both first moment of image and basis images. First moment which is a method for finding centroid of image, is applied to exclude the needless backgrounds in the face recognitions by shifting to the centroid of face image. Basis images which are the face features, are respectively extracted by principal component analysis(PCA) and fixed-point independent component analysis(FP-ICA). This is to improve the recognition performance by excluding the redundancy considering to second- and higher-order statistics of face image. The proposed methods has been applied to the problem for recognizing the 48 face images(12 persons*4 scenes) of 64*64 pixels. The 3 distances such as city-block, Euclidean, negative angle are used as measures when match the probe images to the nearest gallery images. The experimental results show that the proposed methods has a superior recognition performances(speed, rate) than conventional PCA and FP-ICA without preprocessing, the proposed FP-ICA has also better performance than the proposed PCA. The city-block has been relatively achieved more an accurate similarity than Euclidean or negative angle.

• The Journal of the Korea Contents Association
• /
• v.3 no.2
• /
• pp.19-23
• /
• 2003

Neurons in the mammalian visual cortex can be dassified into the two main categories of simple cells and complex cells based on their response properties. Here, we find the complex features corresponding to the response of complex cells by applying the unsupervised independent component analysis network to input images. This result will be helpful to elucidate the information processing mechanism of neurons in primary visual cortex.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.679-688
• /
• 2021

Gait is an evaluation index used in various clinical area including brain nervous system diseases. Signal source localizing and time-frequency analysis are mainly used after extracting independent components for Electroencephalogram data as a method of measuring and analyzing brain activation related to gait. Existing treadmill-based walking EEG analysis performs signal preprocessing, independent component analysis(ICA), and source localizing by merging data after the multiple EEG measurements, and extracts representative component clusters through inter-subject clustering. In this study we propose an analysis method, without merging to single dataset, that performs signal preprocessing, ICA, and source localization on each measurements, and inter-subject clustering is conducted for ICs extracted from all subjects. The effect of data merging on the IC clustering and time-frequency analysis was investigated for the proposed method and two conventional methods. As a result, it was confirmed that a more subdivided gait-related brain signal component was derived from the proposed "non-merging" method (4 clusters) despite the small number of subjects, than conventional method (2 clusters).