Unconstrained Object Segmentation Using GrabCut Based on Automatic Generation of Initial Boundary

Na, In-Seop,Oh, Kang-Han,Kim, Soo-Hyung The Korea Contents Association 2013 International Journal of Contents Vol.9 No.1

Foreground estimation in object segmentation has been an important issue for last few decades. In this paper we propose a GrabCut based automatic foreground estimation method using block clustering. GrabCut is one of popular algorithms for image segmentation in 2D image. However GrabCut is semi-automatic algorithm. So it requires the user input a rough boundary for foreground and background. Typically, the user draws a rectangle around the object of interest manually. The goal of proposed method is to generate an initial rectangle automatically. In order to create initial rectangle, we use Gabor filter and Saliency map and then we use 4 features (amount of area, variance, amount of class with boundary area, amount of class with saliency map) to categorize foreground and background. From the experimental results, our proposed algorithm can achieve satisfactory accuracy in object segmentation without any prior information by the user.

Adaptive MCMC-Based Particle Filter for Real-Time Multi-Face Tracking on Mobile Platforms

Na, In Seop,Le, Ha,Kim, Soo Hyung The Korea Contents Association 2014 International Journal of Contents Vol.10 No.3

In this paper, we describe an adaptive Markov chain Monte Carlo-based particle filter that effectively addresses real-time multi-face tracking on mobile platforms. Because traditional approaches based on a particle filter require an enormous number of particles, the processing time is high. This is a serious issue, especially on low performance devices such as mobile phones. To resolve this problem, we developed a tracker that includes a more sophisticated likelihood model to reduce the number of particles and maintain the identity of the tracked faces. In our proposed tracker, the number of particles is adjusted during the sampling process using an adaptive sampling scheme. The adaptive sampling scheme is designed based on the average acceptance ratio of sampled particles of each face. Moreover, a likelihood model based on color information is combined with corner features to improve the accuracy of the sample measurement. The proposed tracker applied on various videos confirmed a significant decrease in processing time compared to traditional approaches.

Automatic Segmentation of Product Bottle Label Based on GrabCut Algorithm

Na, In Seop,Chen, Yan Juan,Kim, Soo Hyung The Korea Contents Association 2014 International Journal of Contents Vol.10 No.4

In this paper, we propose a method to build an accurate initial trimap for the GrabCut algorithm without the need for human interaction. First, we identify a rough candidate for the label region of a bottle by applying a saliency map to find a salient area from the image. Then, the Hough Transformation method is used to detect the left and right borders of the label region, and the k-means algorithm is used to localize the upper and lower borders of the label of the bottle. These four borders are used to build an initial trimap for the GrabCut method. Finally, GrabCut segments accurate regions for the label. The experimental results for 130 wine bottle images demonstrated that the saliency map extracted a rough label region with an accuracy of 97.69% while also removing the complex background. The Hough transform and projection method accurately drew the outline of the label from the saliency area, and then the outline was used to build an initial trimap for GrabCut. Finally, the GrabCut algorithm successfully segmented the bottle label with an average accuracy of 92.31%. Therefore, we believe that our method is suitable for product label recognition systems that automatically segment product labels. Although our method achieved encouraging results, it has some limitations in that unreliable results are produced under conditions with varying illumination and reflections. Therefore, we are in the process of developing preprocessing algorithms to improve the proposed method to take into account variations in illumination and reflections.

Automatic Segmentation of Product Bottle Label Based on GrabCut Algorithm

In Seop Na,Yan Juan Chen,Soo Hyung Kim 한국콘텐츠학회(IJOC) 2014 International Journal of Contents Vol.10 No.4

In this paper, we propose a method to build an accurate initial trimap for the GrabCut algorithm without the need for human interaction. First, we identify a rough candidate for the label region of a bottle by applying a saliency map to find a salient area from the image. Then, the Hough Transformation method is used to detect the left and right borders of the label region, and the k-means algorithm is used to localize the upper and lower borders of the label of the bottle. These four borders are used to build an initial trimap for the GrabCut method. Finally, GrabCut segments accurate regions for the label. The experimental results for 130 wine bottle images demonstrated that the saliency map extracted a rough label region with an accuracy of 97.69% while also removing the complex background. The Hough transform and projection method accurately drew the outline of the label from the saliency area, and then the outline was used to build an initial trimap for GrabCut. Finally, the GrabCut algorithm successfully segmented the bottle label with an average accuracy of 92.31%. Therefore, we believe that our method is suitable for product label recognition systems that automatically segment product labels. Although our method achieved encouraging results, it has some limitations in that unreliable results are produced under conditions with varying illumination and reflections. Therefore, we are in the process of developing preprocessing algorithms to improve the proposed method to take into account variations in illumination and reflections.

Genotyping of avian pathogenic Escherichia coli by DNA fragment analysis for the differences in simple sequence repeats

Mi Na Han,Hyeon Seop Byeon,Seong Tae Han,Rae Hoon Jang,Chang Seop Kim,최석화 한국동물위생학회 2018 한국동물위생학회지 (KOJVS) Vol.41 No.4

Avian pathogenic E. coli (APEC) causes severe economic losses in the poultry farms, due to systemic infections leading to lethal colisepticemia. It causes a variety of diseases from air sac infection to sys-temic spread leading to septicemia. Secondary infection contains opportunistic infections due to im-munosuppression disease. Collibacillosis causes the great problems in the poultry industry in Korea. Thus, it is necessary to identify and classify the characteristics of E. coli isolate of chicken origin to confirm the diversity of symptoms and whether they are transmitted among the farms. Fragment analysis is identify the difference in the number of Variable-Number Tandem-Repeats (VNTRs) for genotyping. VNTRs have repeating structure (Microsatellite, Short tandem repeats; STR, Simple sequence repeats; SSR) in the chromosome. This region can be used as a genetic marker because of its high mutation rate. And various lengths of the amplified DNA fragment cause the difference in the number of repeti-tion of the DNA specific site. The number of repetition sequences indicates the separated size of frag-ments, so the each fragments can be distinguished by specific samples. The results of the sample show that there is no difference in six microsatellite loci (yjiD, aidB, molR_1, ftsZ, b1668, yibA). There are differences among the farms in relation of the number of repetitions of other six microsatellite loci (ycgW, yaiN, yiaB, mhpR, b0829, caiF). Four (ycgW, yiaB, b0829, caiF) of these six microsatellite loci show statistically significant differences (P<0.05). It means that the analysis using four microsatellite loci including ycgW, yiaB, b0829, and caiF can confirm among the farms. Five E. coli samples in one farm have same SSR repetition at all markers. But, there are significant differences from other farms at Four (ycgW, yiaB, b0829, caiF) microsatellite loci. These results emphasize again that the four micro-satellite loci makes a difference in the amplified DNA fragments, enabling it to be used for E. coli genotyping.

Adaptive MCMC-Based Particle Filter for Real-Time Multi-Face Tracking on Mobile Platforms

In Seop Na,Ha Le,Soo Hyung Kim 한국콘텐츠학회 2014 International Journal of Contents Vol.10 No.3

Magnetic Resonance Imaging of Ferumoxytol-Labeled Human Mesenchymal Stem Cells in the Mouse Brain

Lee, Na Kyung,Kim, Hyeong Seop,Yoo, Dongkyeom,Hwang, Jung Won,Choi, Soo Jin,Oh, Wonil,Chang, Jong Wook,Na, Duk L. Springer US 2017 Stem cell reviews and reports Vol.13 No.1

<P>The success of stem cell therapy is highly dependent on accurate delivery of stem cells to the target site of interest. Possible ways to track the distribution of MSCs in vivo include the use of reporter genes or nanoparticles. The U.S. Food and Drug Administration (FDA) has approved ferumoxytol (Feraheme® [USA], Rienso® [UK]) as a treatment for iron deficiency anemia. Ferumoxytol is an ultrasmall superparamagnetic iron oxide nanoparticle (USPIO) that has recently been used to track the fate of transplanted cells using magnetic resonance imaging (MRI). The major objectives of this study were to demonstrate the feasibility of labeling hUCB-MSCs with ferumoxytol and to observe, through MRI, the engraftment of ferumoxytol-labeled human umbilical cord blood-derived mesenchymal stem cells (hUCB-MSCs) delivered via stereotactic injection into the hippocampi of a transgenic mouse model of familial Alzheimer’s disease (5XFAD). Ferumoxytol had no toxic effects on the viability or stemness of hUCB-MSCs when assessed <I>in vitro</I>. Through MRI, hypointense signals were discernible at the site where ferumoxytol-labeled human MSCs were injected. Iron-positive areas were also observed in the engrafted hippocampi. The results from this study support the use of nanoparticle labeling to monitor transplanted MSCs in real time as a follow-up for AD stem cell therapy in the clinical field.</P><P><B>Electronic supplementary material</B></P><P>The online version of this article (doi:10.1007/s12015-016-9694-0) contains supplementary material, which is available to authorized users.</P>

Co-infection of Aspergillus spp. with Erysipelothrix rhusiopathiae in a red-crowned crane: a case report

Mi-Na Han,Jeong-Ho Kim,Sang-Kab Lee,Kyong-Ok No,Mun-Hui Chae,Woo-Kyoung Cho,Jong-Hwa Lee,Rae-Hoon Jang,Chang-Seop Kim,Hyeon-Seop Byeon 한국동물위생학회 2020 한국동물위생학회지 (KOJVS) Vol.43 No.2

This case describes outbreaks of acute aspergillosis in a red-crowned crane. A six-month-old, male, crane had showed clinical signs (i.e. anorexia, performance loss, ruffled feathers and drooped wings and open mouth breathing, etc.) before death. In necropsy examination, spherical to oval nodules dissemi-nated from the respiratory tract to other organs. Those nodules were formed predominantly in air sacs, lung, peritoneum, serosa of esophagus and trachea. The nodules varied in size from 1 mm to over 1cm and the color was white to yellow. Microscopically, most of lung architecture were replaced by multiple foci which were characterized by well demarcated eosinophilic and karyorrhetic debris and surrounded by numerous Inflammatory cell. Most within necrotic center of the nodules, large numbers of fungal hyphae were present. Microbiology result indicated fungal growths on sabroud dextrose agar and bacte-rial growths on blood agar. Bacteria identified as E. rhusiopathiae using MALDI-TOF (microflex, BRUKER, USA) and fungi identified as A. fumigatus, A. terreus by sequencing the ITS1 and ITS4 regions. To con-firm the route of infection, we checked the existence of the same pathogens in cohabitant (i.e. mother crane). The young age and weakened immunity (i.e. bacterial infection, etc.) causes fatal aspergillosis in birds.

Unconstrained Object Segmentation Using GrabCut Based on Automatic Generation of Initial Boundary

In-Seop Na,Kang-Han Oh,Soo-Hyung Kim 한국콘텐츠학회(IJOC) 2013 International Journal of Contents Vol.9 No.1

Foreground estimation in object segmentation has been an important issue for last few decades. In this paper we propose a GrabCut based automatic foreground estimation method using block clustering. GrabCut is one of popular algorithms for image segmentation in 2D image. However GrabCut is semi-automatic algorithm. So it requires the user input a rough boundary for foreground and background. Typically, the user draws a rectangle around the object of interest manually. The goal of proposed method is to generate an initial rectangle automatically. In order to create initial rectangle, we use Gabor filter and Saliency map and then we use 4 features (amount of area, variance, amount of class with boundary area, amount of class with saliency map) to categorize foreground and background. From the experimental results, our proposed algorithm can achieve satisfactory accuracy in object segmentation without any prior information by the user.

Sequence analysis of the fusion protein gene of Newcastle disease virus isolated from breeder ducks in Korea

Mi Na Han,Hyeon Seop Byeon,Cho Yeon Lee,Nam Sin Jo,Jong Hwa Lee,Rae Hoon Jang,Chang Seop Kim,Ki Jeong Na 한국동물위생학회 2019 한국동물위생학회지 (KOJVS) Vol.42 No.4

Newcastle disease (ND) is an infectious poultry disease that caused high mortality and reduced egg production. NDVs are regularly present in the domestic duck population. And ducks play a possible role in the maintenance and transmission of NDVs. While we were monitoring the Avian Influenza, NDVs were isolated from field samples by accident. So we analysed the biological and genetic characteristics of these viruses. Lentogenic NDVs were isolated from two farms among twenty breeder duck farms. The ages of ducks were 39 weeks old in the ‘A’ farm and 3∼72 weeks old in the ‘B’ farm. And they were not inoculated with the NDVs vaccine. In the biological characteristics, the both viruses which separated from the farm ‘A’ and ‘B’ were thermostable. The amino acid sequence of a site from 112 to 119 in the fusion (F) protein was ‘GKQGRLIG’ which has monobasic motif in the samples of both farms. And this means the separated NDVs are lentogenic. Phylogenetic analysis was performed by en-tire nucleotide sequence of F protein. The virus strains from the A farm (MN095239) and the B farm (MN095240) belonged to class II genotype I. Using the analysis of whole F protein nucleic acid se-quence, the MN095239 (GenBank) had homology with Ulster strain about 99.95% and the MN095239 (GenBank) had homology with KR/CK/KU_LBM255/09 strain about 99.89%. NDV surveillance is needed to investigate epidemiological relationship of domestic breeder duck isolates in Korea.