• Journal of the Korea Society of Computer and Information
• /
• v.19 no.3
• /
• pp.37-43
• /
• 2014

In this paper, we propose a robust lip detection algorithm using color clustering. At first, we adopt AdaBoost algorithm to extract facial region and convert facial region into Lab color space. Because a and b components in Lab color space are known as that they could well express lip color and its complementary color, we use a and b component as the features for color clustering. The nearest neighbour clustering algorithm is applied to separate the skin region from the facial region and K-Means color clustering is applied to extract lip-candidate region. Then geometric characteristics are used to extract final lip region. The proposed algorithm can detect lip region robustly which has been shown by experimental results.

• Korean Journal of Ecology and Environment
• /
• v.36 no.1 s.102
• /
• pp.9-20
• /
• 2003

Response-species of zooplankton on precipitation and temperature were investigat-ed in the tributaries of Han River from May to November, 2002. Total 42 species(Rotifer twenty three species, Cladocera ten species, Copepoda eight species and Protozoa one species) were collected in the target location. The highest abundance and the number of species were observed in May. Due to the summer precipitation the abundance of Rotifer and Cladocera were largely decreased in August. Dominant species of Rotifer appeared Brachionus urceolaris, Euclanis dilatata, Lecane Inua, Brachionus quadridentatus and Brachionus calyciflorus in May, Monostyla bulla and Conochilus unicornis in August and Brachionus quadridentatus, Euclanis dilatata and Lecane Iuna in November. Also, the dominant species of Cladocera were Moina weismanni, Simocephalus vetulus, Scapholebris aucrgnafa, Chydorus sphaericus in May, Chydorus sphaericus, Scapholebris mucronata in August, Chydorus sphaericus and Alona rectangular in November. Nearest neighbour clustering was implemented for classification of zooplankton abundance indifferent month of survey locations. The patterning of May and November appeared similar but August was different.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.3
• /
• pp.1149-1154
• /
• 2014

Background: In Malaysia, data from the Malaysian Health Ministry showed colorectal cancer (CRC) to be the second most common type of cancer in 2007-2009, after breast cancer. The same was apparent after looking at males and females cases separately. In the present study, the Geographic Information System (GIS) was employed to describe the distribution of CRC cases in Kuala Lumpur (KL), Malaysia, according to socio-demographic factors (age, gender, ethnicity and district). Materials and Methods: This retrospective review concerned data for patients diagnosed with colorectal cancer in the years 1995 to 2011 collected from the Wilayah Persekutuan Health Office, taken from the cancer notification form (NCR-2), and patient medical records from the Surgical Department, Universiti Kebangsaan Malaysia Medical Centre (UKMMC). A total of 146 cases were analyzed. All the data collected were analysed using ArcGIS version 10.0 and SPSS version 19.0. Results: Patients aged 60 to 69 years accounted for the highest proportion of cases (34.2%) and males slightly predominated 76 (52.1%), Chinese had the highest number of registered cases at 108 (74.0%) and staging revealed most cases in the 3rd and 4th stages. Kernel density analysis showed more cases are concentrated up in the northern area of Petaling and Kuala Lumpur subdistricts. Spatial global pattern analysis by average nearest neighbour resulted in nearest neighbour ratio of 0.75, with Z-score of -5.59, p value of <0.01 and the z-score of -5.59. Spatial autocorrelation (Moran's I) showed clustering significant with p<0.01, Z score 3.14 and Moran's Index of 0.007. When mapping clusters with hotspot analysis (Getis-Ord Gi), hot and cold spots were identified. Hot spot areas fell on the northeast side of KL. Conclusions: This study demonstrated significant spatial patterns of cancer incidence in KL. Knowledge about these spatial patterns can provide useful information to policymakers in the planning of screening of CRC in the targeted population and improvement of healthcare facilities to provide better treatment for CRC patients.