Browsing by Author "Orawan Chaowalit"

Now showing 1 - 3 of 3
  • Default Image
    Item
    Clustering tourist using DBSCAN algorithm
    (American Institute of Physics Inc., 2022) Fuangfar Pensiri; Porawat Visutsak; Orawan Chaowalit; O. Chaowalit; King Mongkut's University of Technology North Bangkok University, Bangkok, 1518 Pracharat 1 Road Wongsawang Bangsue, 10800, Thailand; email: orawan@su.ac.th
    The tourist clustering refer the aggregating of prospective tourist into different groups with common observance by using statistical data analysis technique. In this paper, we apply the Density-based spatial clustering of applications with noise (DBSCAN) to find the factors that can segment the tourist associated with using digital technology equipment as a tourism facility based on the data of tourist behaviour and activity. We describe the methodology, firstly analyse the algorithm. Secondly, compare execution of the different parameter values (Eps): the maximum radius of the neighbourhood from core point and the minimum number of points required to form a dense region (MinPts). Finally, examine the outcome of the application, Tourist's career and tourism style are two factors from eleven factors can cluster the tourists into eight groups with Eps and MinPts parameters 0.5 and 10 respectively. © 2022 Author(s).
  • Default Image
    Item
    Image Analysis of Mushroom Types Classification by Convolution Neural Networks
    (Association for Computing Machinery, 2019) Jitdumrong Preechasuk; Orawan Chaowalit; Fuangfar Pensiri; Porawat Visutsak
    Mushrooms are fungi. The edible mushrooms include nutritional content and health benefits. However, some mushroom species is toxic and contains poisonous substances that could cause illness and lead to death. Mushroom poisoning accounts for approximately 70% of natural poisoning and often causes death. However, there are only 30-50 poisonous species among the thousands of species found on earth, and of these, no more than 10 are fatally poisonous [1]. The main reason of eating poisonous mushrooms is the lack of knowledge and skill to classify the edible and poisonous mushrooms. Besides, the physical characteristics of mushrooms are similar. Therefore, this work focuses on the classification of 45 types of mushrooms. This work aims to reduce the number of illness persons whom are risk of exposure to toxic mushrooms. This work proposes a new model of classifying 45 types of mushrooms including edible and poisonous mushrooms by using a technique of Convolution Neural Networks. The proposed model was tested on both types of mushroom. It was trained to construct the CNN models and used the trained models to classify all types of mushroom. The proposed model gives the results of 0.78, 0.73 and 0.74 of precision, recall and F1 score, respectively. It concluded that the proposed model can classify types of mushroom image with efficiently and effectively. � 2019 ACM.
  • Default Image
    Item
    Smooth voxel surface for medical volumetric rendering
    (SPIE, 2019) Porawat Visutsak; Fuangfar Pensiri; Orawan Chaowalit
    This paper aims to implement the trilinear interpolation algorithm with the marching cubes method for generating the smooth voxel surface from 2D digital images. The trilinear interpolation is a straight extension of the bilinear interpolation technique. It can be seen as the linear interpolation of two bilinear interpolations. The novel method is a fast and easy to implement and it also produces a smooth results (compared to the marching cubes technique). Therefore, for volume rendering such as the 3D medical models and terrains where a very large numbers of lookups in 3D grids are performed, this method is a very good choice for generating the high resolution of 3D surfaces. � 2019 SPIE.