• ETRI Journal
• /
• v.32 no.5
• /
• pp.784-794
• /
• 2010

Automatic facial expression recognition has many potential applications in different areas of human computer interaction. However, they are not yet fully realized due to the lack of an effective facial feature descriptor. In this paper, we present a new appearance-based feature descriptor, the local directional pattern (LDP), to represent facial geometry and analyze its performance in expression recognition. An LDP feature is obtained by computing the edge response values in 8 directions at each pixel and encoding them into an 8 bit binary number using the relative strength of these edge responses. The LDP descriptor, a distribution of LDP codes within an image or image patch, is used to describe each expression image. The effectiveness of dimensionality reduction techniques, such as principal component analysis and AdaBoost, is also analyzed in terms of computational cost saving and classification accuracy. Two well-known machine learning methods, template matching and support vector machine, are used for classification using the Cohn-Kanade and Japanese female facial expression databases. Better classification accuracy shows the superiority of LDP descriptor against other appearance-based feature descriptors.

• The KIPS Transactions:PartD
• /
• v.13D no.1 s.104
• /
• pp.29-36
• /
• 2006

In this paper, we present a leaf image retrieval system that represents and retrieves leaf images based on their shape. For more effective representation of leaf images, we improved an existing MPP algorithm. Also, in order to reduce the response time, we proposed a new dynamic matching algorithm at basically revises the Nearest Neighbor search. The system provides users with an interface for uploading query images or tools to generate queries based on shape features and retrieves images based on their similarity. For convenience, users are allowed to easily query images by sketching leaf shape or leaf arrangement on the web. In the experiment, we constructed an image database of Korean native plants and measured the system performance by counting the number of similar images retrieved for queries.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.3
• /
• pp.385-392
• /
• 2006

This paper proposes a new stroke extraction algorithm for phoneme segmentation, which is one of main techniques for off-line handwritten Hangul recognition. The proposed algorithm extracts vertical, slant, and horizontal strokes from phonemes using run-length. The run-length of vertical or slant strokes becomes the width, and also the number of horizontal run-lengths the width. After extracting horizontal strokes from phonemes, the algorithm links two continuous vertical or slant stokes with run-lengths of the strokes' width to represent the features of a character. The extracted strokes can be utilized to recognize a character, using template matching of strokes, which is being adopted in on-line handwritten Hangul recognition.

• Journal of Broadcast Engineering
• /
• v.14 no.4
• /
• pp.397-408
• /
• 2009

In this paper, we present a non-static color histogram method to retrieve clothing images that are similar to a query clothing. Given clothing area, our method automatically extracts major colors by using the octree-based quantization approach[16]. Then, a color palette that is composed of the major colors is generated. The feature of each clothing, which can be either a query or a database clothing image, is represented as a color histogram based on its color palette. We define the match color bins between two possibly different color palettes, and unify the color palettes by merging or deleting some color bins if necessary. The similarity between two histograms is measured by using the weighted Euclidean distance between the match color bins, where the weight is derived from the frequency of each bin. We compare our method with previous histogram matching methods through experiments. Compared to HSV cumulative histogram-based approach, our method improves the retrieval precision by 13.7 % with less number of color bins.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.1
• /
• pp.38-57
• /
• 2016

Service discovery is one of the major challenges in cloud computing environment with a large number of service providers and heterogeneous services. Non-uniform naming conventions, varied types and features of services make cloud service discovery a grueling problem. With the proliferation of cloud services, it has been laborious to find services, especially from Internet-based service repositories. To address this issue, services are crawled and clustered according to their similarity. The clustered services are maintained as a catalogue in which the data published on the cloud provider's website are stored in a standard format. As there is no standard specification and a description language for cloud services, new efficient and intelligent mechanisms to discover cloud services are strongly required and desired. This paper also proposes a key-value representation to describe cloud services in a formal way and to facilitate matching between offered services and demand. Since naïve users prefer to have a query in natural language, semantic approaches are used to close the gap between the ambiguous user requirements and the service specifications. Experimental evaluation measured in terms of precision and recall of retrieved services shows that the proposed approach outperforms existing methods.

• Journal of Microbiology and Biotechnology
• /
• v.7 no.6
• /
• pp.397-401
• /
• 1997

The present research program was conducted to characterize a strain of actinomycetes producing an anti methicillin-resistant Staphylococcus aureus (MRSA) antibiotic. Soil samples were collected from various sites in Korea and a number of actinomycetes were isolated from the soil samples by applying selective agar for actinomycetes. Among over 400 isolates, a strain (AMLK-135) producing anti-MRSA antibiotic against S. aureus TK 784 was selected. According to the morphological and physiological characteristics, the strain AMLK-135 was confirmed to belong to the genus Streptomyces. From the results of species identification with the TAXON program, the strain AMLK-135 was shown to belong to major cluster 5 (Streptomyces exfoliatus), but it had a low simple matching coefficient ($S_{SM}$ SM/) value to member organisms of major cluster 5. Percentage ($\%$) of strain further away of the strain AMLK-135 was low (1.9400) and it was placed further away than the outer-most members in major cluster 5. Therefore, the strain AMLK-135 was identified as a new species of the genus Streptomyces.

• Journal of Korea Multimedia Society
• /
• v.20 no.12
• /
• pp.1865-1873
• /
• 2017

The feature-based visual SLAM requires 3D positions for the extracted feature points to perform 3D-2D motion estimation. LiDAR can provide reliable and accurate 3D position information with low computational burden, while stereo camera has the problem of the impossibility of stereo matching in simple texture image region, the inaccuracy in depth value due to error contained in intrinsic and extrinsic camera parameter, and the limited number of depth value restricted by permissible stereo disparity. However, the sparsity of LiDAR data may increase the inaccuracy of motion estimation and can even lead to the result of motion estimation failure. Therefore, in this paper, we propose three interpolation methods which can be applied to interpolate sparse LiDAR data. Simulation results obtained by applying these three methods to a visual odometry algorithm demonstrates that the selective bilinear interpolation shows better performance in the view point of computation speed and accuracy.

• Journal of Advanced Navigation Technology
• /
• v.6 no.1
• /
• pp.27-36
• /
• 2002

As the number of autonomous vehicles is increased, drivers are trending toward constant interests in detecting distance between vehicles in close-range and maintaining the distance between forward and backward vehicles for drivers' safety. In this paper, we propose a method of detecting the distance between two vehicles by computing the disparity of the close-rang vehicle using stereo vision. The boundary of the vehicle is obtained by using the modified wavelet transform which has multi-resolution characteristics. Then the disparity between left and right images is computed using coarse-to-fine method and histogram matching. Here we transform the left-right stereo images through 3-steps using the modified wavelet for maintaining the original resolution. An experimental result showed that the proposed method had 4.65% in total error rate.

• Journal of KIISE:Databases
• /
• v.31 no.4
• /
• pp.341-351
• /
• 2004

With the popularity of digital audio contents, querying and retrieving audio contents efficiently from database has become essential. In this paper, we propose a new index scheme for retrieving audio contents efficiently using audio portions that have been queried frequently. This scheme is based on the observation that users have a tendency to memorize and query a small number of audio portions. Detecting and indexing such portions enables fast retrieval and shows better performance than sequential search-based audio retrieval. Moreover, this scheme is independent of underlying retrieval system, which means this scheme can work together with any other audio retrieval system. We have implemented a prototype system and showed its performance gain through experiments.

• Korean Journal of Veterinary Research
• /
• v.43 no.1
• /
• pp.95-102
• /
• 2003

Because Staphylococcus aureus (S. aureus) has variable number of short sequence repeat region in coagulase gene, it has been used to investigate the relatedness of S. aureus isolates. In this study, we isolated S. aureus strains from 20 dairy farms with bovine mastitis from September 2000 to August 2001. PCR-RFLP analysis of coagulase gene revealed 10 different patterns. Most of the S. aureus isolates showed only one coagulase gene RFLP pattern per farm. However, there were several S. aureus clones spreading between dairy farms. All the farms showed poor management conditions of milking machine and milker, indicating that managements for mastitis control program include use of proper milking matching, premilking sanitation, and segregation in the S. aureus infection herd. Our data suggest that PCR-RFLP analysis of coagulase gene might be applicable for the epidemiological investigations of S. aureus isolated from bovine mastitis cows.