• Journal of Information Processing Systems
• /
• v.12 no.3
• /
• pp.468-488
• /
• 2016

The paper proposes a novel framework for 3D face verification using dimensionality reduction based on highly distinctive local features in the presence of illumination and expression variations. The histograms of efficient local descriptors are used to represent distinctively the facial images. For this purpose, different local descriptors are evaluated, Local Binary Patterns (LBP), Three-Patch Local Binary Patterns (TPLBP), Four-Patch Local Binary Patterns (FPLBP), Binarized Statistical Image Features (BSIF) and Local Phase Quantization (LPQ). Furthermore, experiments on the combinations of the four local descriptors at feature level using simply histograms concatenation are provided. The performance of the proposed approach is evaluated with different dimensionality reduction algorithms: Principal Component Analysis (PCA), Orthogonal Locality Preserving Projection (OLPP) and the combined PCA+EFM (Enhanced Fisher linear discriminate Model). Finally, multi-class Support Vector Machine (SVM) is used as a classifier to carry out the verification between imposters and customers. The proposed method has been tested on CASIA-3D face database and the experimental results show that our method achieves a high verification performance.

• The Korean Journal of Applied Statistics
• /
• v.24 no.1
• /
• pp.11-23
• /
• 2011

Robust unit root tests are developed for dynamic panels consisting of TAR processes. The test statistics are all based on diverse combinations of individual t-type tests for significance of TAR coefficients. Limiting null distributions are established. A Monte-Carlo experiment compares the proposed tests. The tests are applied to a panel data set of Canadian unemployment rates which show asymmetric features as well as having outliers.

• Journal of information and communication convergence engineering
• /
• v.5 no.1
• /
• pp.78-82
• /
• 2007

This paper propose a building reconstruction method of urban area for a 3D GIS with stereo images. The 3D reconstruction is performed by the grouping 3D line segments extracted from the stereo matching of salient edges which are derived from multiple images. The grouping is achieved by conditions of degrees and distances between lines. Building objects are determined by the junction combinations of the grouped line segments. The proposed algorithm demonstrates effective results of 3D reconstruction of buildings with 2D aerial images.

• Parasites, Hosts and Diseases
• /
• v.51 no.4
• /
• pp.393-400
• /
• 2013

Ocular toxoplasmosis is a disease caused by the infection with Toxoplasma gondii through congenital or acquired routes. Once the parasite reaches the retina, it proliferates within host cells followed by rupture of the host cells and invasion into neighboring cells to make primary lesions. Sometimes the restricted parasite by the host immunity in the first scar is activated to infect another lesion nearby the scar. Blurred vision is the main complaint of ocular toxoplasmic patients and can be diagnosed by detection of antibodies or parasite DNA. Ocular toxoplasmosis needs therapy with several combinations of drugs to eliminate the parasite and accompanying inflammation; if not treated it sometimes leads to loss of vision. We describe here clinical features and currently available chemotherapy of ocular toxoplasmosis.

• Journal of Mechanical Science and Technology
• /
• v.15 no.2
• /
• pp.217-224
• /
• 2001

In this paper, the relocation of a mobile robot is considered such that it enables the robot to determine its position with respect to a global reference frame without any $\alpha$ priori position information. The robot acquires sonar range data from a two-dimensional model composed of planes, corners, edges, and cylinders. Considering individual range as data features, the robot searches the best position where the data features of a position matches the environmental model using a constraint-based search method. To increase the search efficiency, a hypothesize and-verify technique is employed in which the position of the robot is calculated from all possible combinations of two range returns that satisfy the sonar sensing model. Accurate relocation is demonstrated with the results from sets of experiments using sparse sonar data in the presence of unmodeled objects.

• Journal of Advanced Navigation Technology
• /
• v.28 no.2
• /
• pp.238-245
• /
• 2024

In this paper, we propose a framework for improving the performance of semantic segmentation of agricultural multispectral image using feature fusion techniques. Most of the semantic segmentation models being studied in the field of smart farms are trained on RGB images and focus on increasing the depth and complexity of the model to improve performance. In this study, we go beyond the conventional approach and optimize and design a model with multispectral and attention mechanisms. The proposed method fuses features from multiple channels collected from a UAV along with a single RGB image to increase feature extraction performance and recognize complementary features to increase the learning effect. We study the model structure to focus on feature fusion and compare its performance with other models by experimenting with favorable channels and combinations for crop images. The experimental results show that the model combining RGB and NDVI performs better than combinations with other channels.

• The Korean Journal of Cytopathology
• /
• v.2 no.2
• /
• pp.90-97
• /
• 1991

Liver is generally known as an organ which is most commonly involved by the metastic tumors. According to the tendency of using fine needle aspiration in the diagnosis of hepatic tumors, the differentital diagnosis between hepatocellular carcinoma and metastatic carcinoma frequently has been a main issue in the poorly differentitated cases, especially to the pathologists of Korea, an endemic area of hepatocellular carcinoma. Until now the problem has been usually solved by the comparison of cytologic characteristics of their tumor cells but not by background cytologic features which rarely have been studied. We observed the background cytologic features helpful for the differential diagnosis through the analysis of 20 cases who had confirmed primary cancer and were diagnosed as metastatic carcinomas in the liver by fine needle aspiration cytology. Twenty cases included 9 adenocarcinomas, 7 spuamous cell carcinomas, 1 small cell carcinoma, 1 carcinoid, 1 adenoid cystic carcinoma, and 1 renal cell cacinoma. Analysis of background cytologic features revealed that 77% of adenocacinoma cases showed benign mesenchymal components and hepatocytes and spuamous cell carcinoma cases disclosed benign mesenchymal tissue (71%) and necrosis (57%), Remaining cases showed variable combinations of benign mesenchymal component, necrosis, hepatocytes, and bile duct epithelial cells. No case revealed atypical hepatocytic naked nuclei, a useful cytologic finding of hepatocellular carcinoma. In summary, the background cytologic features more commonly observed in metastatic carcinomas than in the hepatocellular carcinoma were benign mesenchymal components, hepatocytes, necrosis, and bile duct epithelium. The endothelial cells and hepatocytic naked nuclei, two relatively specific findings of hepatocellular carcinoma were not observed except for renal ceil carcinoma. Above background cytologic features are thought to be helpful for the differential diagnosis between the hepatocellular carcinoma and various metastatic carcinomas in the poorly differentiated cases.

• ALGAE
• /
• v.27 no.4
• /
• pp.249-258
• /
• 2012

Nongeniculate coralline red algae are a common element of the Brazilian coastal zone, especially associated to coral reefs. During the course of ecological studies at Parque Municipal Marinho do Recife de Fora, two species of non-geniculate Corallinales were the major organisms covering the reef. Analyses of the vegetative and reproductive features of the species were analyzed; indicating that one new species of the genus Lithothamnion is proposed here based on the combination of several features associated with anatomy of the tetrasporangial conceptacles in relation to other species of the genus for which modern accounts are available. This new proposal along with other new species, new combinations and range extension of some species of the genus based in similar features clearly suggest that stability in species delimitation is possible. The second species found Pneophyllum conicum represents a range extension of more than 6,000 km from the Pacific in to the Atlantic Ocean suggesting that some nongeniculate species are widely distributed. The occurrence and abundance of these species supports and emphasizes the need for an extensive taxonomic reassessment of coralline red algae in the context of Brazilian coral reef biodiversity.

• Proceedings of the KIEE Conference
• /
• 1998.07b
• /
• pp.552-554
• /
• 1998

In target tracking problems the fixed gain Kalman filter is primarily used to predict a target state vector. This filter, however, has a poor precision for maneuvering targets while it has a good performance for non-maneuvering targets. To overcome the problem this paper proposes the system which estimates the acceleration with neural networks using the input estimation technique. The ability to efficiently fuse information of different forms is one of the major capabilities of trained multi-layer neural networks. The primary motivation for employing neural networks in these applications comes from the efficiency with which more features can be utilized as inputs for estimating target maneuvers. The parallel processing capability of a properly trained neural network can permit fast processing of features to yield correct acceleration estimates. The features used as inputs can be extracted from the combinations of innovation data and heading changes, and for this we set the two dimensional model. The properly trained neural network system outputs the acceleration estimates and compensates for the primary Kalman filter. Finally the proposed system shows the optimum performance.

• Korean Journal of Remote Sensing
• /
• v.19 no.5
• /
• pp.351-361
• /
• 2003

As the launches of a series of remote sensing satellites, there are various multiresolution and multi-spectral images available nowadays. This diversity in remotely sensed image data has created a need to be able to integrate data from different sources. The C-band imaging radar of ERS-2 due to its high sensitivity to coastal wetlands holds tremendous potential in mapping and monitoring coastal wetland features. This paper investigates the advantages of using ERS-2 SAR data combined with IRS-ID LISS-3 data for mapping complex coastal wetland features of Tamil Nadu, southern India. We present a methodology in this paper that highlights the mapping potential of different combinations of filtering and integration techniques. The methodology adopted here consists of three major steps as following: (i) speckle noise reduction by comparative performance of different filtering algorithms, (ii) geometric rectification and coregistration, and (iii) application of different integration techniques. The results obtained from the analysis of optical and microwave image data have proved their potential use in improving interpretability of different coastal wetland features of southern India. Based visual and statistical analyzes, this study suggests that brovey transform will perform well in terms of preserving spatial and spectral content of the original image data. It was also realized that speckle filtering is very important before fusing optical and microwave data for mapping coastal mangrove wetland ecosystem.