• Journal of Animal Science and Technology
• /
• v.66 no.1
• /
• pp.167-177
• /
• 2024

Iris segmentation is an initial step for identifying the biometrics of animals when establishing a traceability system for livestock. In this study, we propose a deep learning framework for pixel-wise segmentation of bovine iris with a minimized use of annotation labels utilizing the BovineAAEyes80 public dataset. The proposed image segmentation framework encompasses data collection, data preparation, data augmentation selection, training of 15 deep neural network (DNN) models with varying encoder backbones and segmentation decoder DNNs, and evaluation of the models using multiple metrics and graphical segmentation results. This framework aims to provide comprehensive and in-depth information on each model's training and testing outcomes to optimize bovine iris segmentation performance. In the experiment, U-Net with a VGG16 backbone was identified as the optimal combination of encoder and decoder models for the dataset, achieving an accuracy and dice coefficient score of 99.50% and 98.35%, respectively. Notably, the selected model accurately segmented even corrupted images without proper annotation data. This study contributes to the advancement of iris segmentation and the establishment of a reliable DNN training framework.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.4
• /
• pp.1904-1926
• /
• 2016

Iris recognition for biometric personnel identification has gained much interest owing to the increasing concern with security today. The image quality plays a major role in the performance of iris recognition systems. When capturing an iris image under uncontrolled conditions and dealing with non-cooperative people, the chance of getting non-ideal images is very high owing to poor focus, off-angle, noise, motion blur, occlusion of eyelashes and eyelids, and wearing glasses. In order to improve the accuracy of iris recognition while dealing with non-ideal iris images, we propose a novel algorithm that improves the quality of degraded iris images. First, the iris image is localized properly to obtain accurate iris boundary detection, and then the iris image is normalized to obtain a fixed size. Second, the valid region (iris region) is extracted from the segmented iris image to obtain only the iris region. Third, to get a well-distributed texture image, bilinear interpolation is used on the segmented valid iris gray image. Using contrast-limited adaptive histogram equalization (CLAHE) enhances the low contrast of the resulting interpolated image. The results of CLAHE are further improved by stretching the maximum and minimum values to 0-255 by using histogram-stretching technique. The gray texture information is extracted by 1D Gabor filters while the Hamming distance technique is chosen as a metric for recognition. The NICE-II training dataset taken from UBRIS.v2 was used for the experiment. Results of the proposed method outperformed other methods in terms of equal error rate (EER).

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.7
• /
• pp.9-15
• /
• 2009

The fuzzy weighted mean classifier is one of the most common classification models and could achieve high performance by adjusting the weights. However, the weights were generally decided based on the experience of experts, which made the resulting classifiers to suffer the lack of consistency and objectivity. To resolve this problem, in this paper, a weight deciding method based on the statistics of the data is introduced, which ensures the learned classifiers to be consistent and objective. To investigate the effectiveness of the proposed methods, Iris data set available from UCI machine learning repository is used and promising results are obtained.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.5
• /
• pp.175-182
• /
• 2010

The paper presents a novel algorithm for face and iris detection with the application for driver iris monitoring. The proposed algorithm consists of the following major steps: Skin-color segmentation, facial features segmentation, and iris positioning. For the skin-segmentation we applied a multi-layer perceptron to approximate the statistical probability of certain skin-colors, and filter out those with low probabilities. The next step segments the face region into the following categories: eye, mouth, eye brow, and remaining facial regions. For this purpose we propose a novel segmentation technique based on estimation of facial class probability density functions (PDF). Each facial class PDF is estimated on the basis of salient features extracted from a corresponding facial image region. Then pixels are classified according to the highest probability selected from four estimated PDFs. The final step applies the circular Hough transform to the detected eye regions to extract the position and radius of the iris. We tested our system on two data sets. The first one is obtained from the Web and contains faces under different illuminations. The second dataset was collected by us. It contains images obtained from video sequences recorded by a CCD camera while a driver was driving a car. The experimental results are presented, showing high detection rates.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.2029-2032
• /
• 2004

The quantity of data were rapidly increased recently and caused the data overwhelming. This led to be difficult in searching the required data. The method of eliminating redundant data was needed. One of the efficient methods was Knowledge Discovery in Database (KDD). Generally data can be separate into 2 cases, continuous data and discrete data. This paper describes algorithm that transforms continuous attributes into discrete ones. We present an Improved Class Attribute Interdependence Maximization (ICAIM), which designed to work with supervised data, for discretized process. The algorithm does not require user to predefine the number of intervals. ICAIM improved CAIM by using significant test to determine which interval should be merged to one interval. Our goal is to generate a minimal number of discrete intervals and improve accuracy for classified class. We used iris plant dataset (IRIS) to test this algorithm compare with CAIM algorithm.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.6
• /
• pp.814-820
• /
• 2008

Recently, researches on ontology-based pattern classification have been tried out in many fields. However, in most of the researches, the ontology which represents the knowledge about pattern classification is just referred during the processes of the pattern classification. In this paper, we propose ontology-based fuzzy classifier for pattern classification which is extended from the fuzzy rule-based classifier In order to realize the proposed classifier, we construct an ontology by conceptualizing the method of fuzzy rule-based pattern classification and generate ontology inference rules for pattern classification. Lastly, we show the validity o) the proposed classifier through the experiment of pattern classification on the Fisher's IRIS dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.3
• /
• pp.976-990
• /
• 2020

The k-means algorithm is widely used in academia and industry due to easy and simple implementation, enabling fast learning for complex datasets. However, k-means struggles to classify datasets without prior knowledge of specific domains. We proposed the repulsive k-means (RK-means) algorithm in a previous study to improve the k-means algorithm, using the repulsive force concept, which allows deleting unnecessary cluster centroids. Accordingly, the RK-means enables to classifying of a dataset without domain knowledge. However, three main problems remain. The RK-means algorithm includes a cluster repulsive force offset, for clusters confined in other clusters, which can cause cluster locking; we were unable to prove RK-means provided optimal convergence in the previous study; and RK-means shown better performance only normalize term and weight. Therefore, this paper proposes the advanced RK-means (ARK-means) algorithm to resolve the RK-means problems. We establish an initialization strategy for deploying cluster centroids and define a metric for the ARK-means algorithm. Finally, we redefine the mass and normalize terms to close to the general dataset. We show ARK-means feasibility experimentally using blob and iris datasets. Experiment results verify the proposed ARK-means algorithm provides better performance than k-means, k'-means, and RK-means.

• Journal of Information Processing Systems
• /
• v.16 no.4
• /
• pp.750-759
• /
• 2020

The wide application of various integrated location-based services (LBS social) and tourism application (app) has generated a large amount of trajectory space data. The trajectory data are used to identify popular tourist attractions with high density of tourists, and they are of great significance to smart service and emergency management of scenic spots. A hot spot analysis method is proposed, based on spatial clustering of trajectory stop points. The DBSCAN algorithm is studied with fast clustering speed, noise processing and clustering of arbitrary shapes in space. The shortage of parameters is manually selected, and an improved method is proposed to adaptively determine parameters based on statistical distribution characteristics of data. DBSCAN clustering analysis and contrast experiments are carried out for three different datasets of artificial synthetic two-dimensional dataset, four-dimensional Iris real dataset and scenic track retention point. The experiment results show that the method can automatically generate reasonable clustering division, and it is superior to traditional algorithms such as DBSCAN and k-means. Finally, based on the spatial clustering results of the trajectory stay points, the Getis-Ord Gi* hotspot analysis and mapping are conducted in ArcGIS software. The hot spots of different tourist attractions are classified according to the analysis results, and the distribution of popular scenic spots is determined with the actual heat of the scenic spots.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.685-691
• /
• 2008

The selection of classification boundaries in fuzzy rule- based classification systems is an important and difficult problem. So various methods based on learning processes such as neural network, genetic algorithm, and so on have been proposed for it. In a previous study, we pointed out the limitation of the methods and discussed a method for fuzzy partitioning in the overlapped region on feature space in order to overcome the time-consuming when the additional parameters for tuning fuzzy membership functions are necessary. In this paper, we propose a method to determine three types of classification boundaries(i.e., non-overlapping, overlapping, and a boundary point) on the basis of statistical information of the given dataset without learning by extending the method described in the study. Finally, we show the effectiveness of the proposed method through experimental results applied to pattern classification problems using the modified IRIS and standard IRIS datasets.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.1
• /
• pp.135-142
• /
• 2012

In this paper, we proposed a new architecture called radial basis function-based polynomial neural networks classifier that consists of heterogeneous neural networks such as radial basis function neural networks and polynomial neural networks. The underlying architecture of the proposed model equals to polynomial neural networks(PNNs) while polynomial neurons in PNNs are composed of Fuzzy-c means-based radial basis function neural networks(FCM-based RBFNNs) instead of the conventional polynomial function. We consider PNNs to find the optimal local models and use RBFNNs to cover the high dimensionality problems. Also, in the hidden layer of RBFNNs, FCM algorithm is used to produce some clusters based on the similarity of given dataset. The proposed model depends on some parameters such as the number of input variables in PNNs, the number of clusters and fuzzification coefficient in FCM and polynomial type in RBFNNs. A multiobjective particle swarm optimization using crowding distance (MoPSO-CD) is exploited in order to carry out both structural and parametric optimization of the proposed networks. MoPSO is introduced for not only the performance of model but also complexity and interpretability. The usefulness of the proposed model as a classifier is evaluated with the aid of some benchmark datasets such as iris and liver.