• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1425-1430
• /
• 2018

In this paper, we propose a CNN based deep learning algorithm for semantic segmentation of images. In order to improve the accuracy of semantic segmentation, we combined pixel level object classification and image level object classification. The image level object classification is used to accurately detect the characteristics of an image, and the pixel level object classification is used to indicate which object area is included in each pixel. The proposed network structure consists of three parts in total. A part for extracting the features of the image, a part for outputting the final result in the resolution size of the original image, and a part for performing the image level object classification. Loss functions exist for image level and pixel level classification, respectively. Image-level object classification uses KL-Divergence and pixel level object classification uses cross-entropy. In addition, it combines the layer of the resolution of the network extracting the features and the network of the resolution to secure the position information of the lost feature and the information of the boundary of the object due to the pooling operation.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.216-219
• /
• 2004

In this study, the concept and techniques to generate the level lA, lB and 2A image products have been reviewed. In particular, radiometric and geometric corrections and bands registration used to generate level lA, lB and 2A products have been focused in this study. Radiometric correction is performed to take into account radiometric gain and offset calculated by compensating the detector response non-uniformity. And, in order to compensate satellite altitude, attitude, skew effects, earth rotation and earth curvature, some geometric parameters for geometric corrections are computed and applied. Bands registration process using the matching function between a geometry, which is called 'reference geometry', and another one which is corresponds to the image to be registered is applied to images in case of multi-spectral imaging mode. In order to generate level-lA image products, a simple radiometric processing is applied to a level-0 image. Level-lB image has the same radiometry correction as a level-lA image, but is also issued from some geometric corrections in order to compensate skew effects, Earth rotation effects and spectral misregistration. Level-2A image is generated using some geo-referencing parameters computed by ephemeris data, orbit attitudes and sensor angles. Level lA image is tested by visual analysis. The difference between distances calculated level 1 B image and distances of real coordinate is tested. Level 2A image is tested Using checking points.

• Journal of the Korean Society of Costume
• /
• v.52 no.5
• /
• pp.31-45
• /
• 2002

In this paper. the focus is laid on identifying preferred wedding dress image and its co-relationship with self image of female university students. the biggest potential customer group in the industry. As for the research method. it conducted both review of literature and empirical research method. Through the former approach, four main research questions were derived : 1) What is the preferred wedding dress image of female university students\ulcorner 2) What is the relationship between real self-image and preferred wedding dress image\ulcorner and 3) that between ideal self-image and preferred wedding dress image\ulcorner 4)What is the relationship between the consistency level of the two self-images and preferred wedding dress image\ulcorner In the empirical mode of research, 404 surveys were counted in the final analysis among 450 questionnaires completed by female undergraduate students in Seoul and Chun-an city. Collected data analyzed using factor analysis. frequency analysis. descriptive analysis. scheffe test. multiple-regression analysis and t-test. Results are as follows: first, the sophisticated image was most preferred among female students, followed by elegant splendor. lovable and chaste, feminine and decorative, and characteristic and sexy image. This result indicates how wedding dress trend has a keen sensibility to general fashion trend just like the trend of outfits for everyday life. Secondly, the research results indicated consistent level of co-relationship among the real and ideal self-image and the preference of wedding dress image. And the last the level of consistence between the ideal self-image and the real self-image directly related to the preference level of wedding dress image, showing almost no significance.

• Journal of Information Processing Systems
• /
• v.15 no.1
• /
• pp.7-21
• /
• 2019

In order to improve the edge segmentation effect of the level set image segmentation and avoid the influence of the initial contour on the level set method, a saliency level set image segmentation model based on local Renyi entropy is proposed. Firstly, the saliency map of the original image is extracted by using saliency detection algorithm. And the outline of the saliency map can be used to initialize the level set. Secondly, the local energy and edge energy of the image are obtained by using local Renyi entropy and Canny operator respectively. At the same time, new adaptive weight coefficient and boundary indication function are constructed. Finally, the local binary fitting energy model (LBF) as an external energy term is introduced. In this paper, the contrast experiments are implemented in different image database. The robustness of the proposed model for segmentation of images with intensity inhomogeneity and complicated edges is verified.

• Journal of Korea Multimedia Society
• /
• v.18 no.7
• /
• pp.827-833
• /
• 2015

In this paper, we have compared three level set-based active contour (LSAC) methods on inhomogeneous MR image segmentation which is known as an important role of brain diseases to diagnosis and treatment in early. MR image is often occurred a problem with similar intensities and weak boundaries which have been causing many segmentation methods. However, LSAC method could be able to segment the targets such as the level set based on the local image fitting energy, the local binary fitting energy, and local Gaussian distribution fitting energy. Our implemented and tested the subcortical image segmentations were the corpus callosum and hippocampus and finally demonstrated their effectiveness. Consequently, the level set based on local Gaussian distribution fitting energy has obtained the best model to accurate and robust for the subcortical image segmentation.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.5 no.4
• /
• pp.61-67
• /
• 2009

PACS and Medical Image System use only high level metadata in retrieving desired image nowadays. In order to retrieve Medical Image Data more efficiently, it would be needed to retrieve similarity by utilizing low level metadata as well as keyword retrieval by high level metadata. Thus, In this paper presents that it has realized similarity retrieval by low level metadata on the basis of MPEG-7, and keyword retrieval by high level metadata of DICOM base. It would be also available to look into medical image data in various methods and read accurate image promptly for diagnosis and treatment by retrieval with integrating two metadata.

• Journal of Distribution Science
• /
• v.15 no.1
• /
• pp.71-81
• /
• 2017

Purpose - Consumers heuristically have a specific stereotype on the price level of individual retail format because each format provides them with a different level of purchase satisfaction and emotional benefits. However, if price image which is consumers' overall impression of the aggregate price level of a retailer does not match with their expectations, its price level would be perceived as unfair. It will eventually lead to dissatisfaction and decreased revisit intention. Focused on department store and discount store, this study was designed to verify whether the price fairness plays a role of mediating effect on two influential relationships between price image and post-purchase satisfaction, and price image and repatronage intention. Research design, data, and methodology - A main survey was conducted to 140 students and 128 effective responses were used for the related analysis. T-test, factor analysis, reliability test, and mediated regression analysis were performed. Six hypotheses were developed to examine the mediating effect of price fairness on the two influential relationships between price image and post-purchase satisfaction, and price image and repatronage intention. It was also examined whether the price image of two different retail format is formed differently or not. Results - People perceived the price images of the two retail formats differently. Overall price level of department store is much higher than that of discount store. Analysis results showed that price image did not solely have a significant influence on post-purchase satisfaction unless price fairness as a mediating variable is added. Price fairness turned out to be having a significant influence on relationship between price image and repatronage intention. It influences on repatronage intention directly and also via price fairness. Conclusions - Post-purchase satisfaction can be achieved only if people perceive the price image as fair no matter how the price level is high or low according to traits of retail formats. If they think it's not fair, they would disapprove of the rightness for the price image, and also express their dissatisfaction with it. Consumers willingly make repeated visits to a store if they are convinced of appropriate price level which is perceived as fair, and if they experienced a satisfaction with overall benefits a particular store offered.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.4
• /
• pp.1760-1778
• /
• 2018

For image segmentation with intensity inhomogeneity, many region-based level set methods have been proposed. Some of them however can't get the relatively ideal segmentation results under the severe intensity inhomogeneity and weak edges, and without use of the image gradient information. To improve that, we propose a new level set method combined with local direction gradient in this paper. Firstly, based on two assumptions on intensity inhomogeneity to images, the relationships between segmentation objects and image gradients to local minimum and maximum around a pixel are presented, from which a new pixel classification method based on weight of Euclidian distance is introduced. Secondly, to implement the model, variational level set method combined with image spatial neighborhood information is used, which enhances the anti-noise capacity of the proposed gradient information based model. Thirdly, a new diffusion process with an edge indicator function is incorporated into the level set function to classify the pixels in homogeneous regions of the same segmentation object, and also to make the proposed method more insensitive to initial contours and stable numerical implementation. To verify our proposed method, different testing images including synthetic images, magnetic resonance imaging (MRI) and real-world images are introduced. The image segmentation results demonstrate that our method can deal with the relatively severe intensity inhomogeneity and obtain the comparatively ideal segmentation results efficiently.

• Proceedings of the IEEK Conference
• /
• 2004.08c
• /
• pp.867-871
• /
• 2004

In this paper, a new analytic approach for shape preserving contrast enhancement is presented. Contrast enhancement is achieved by means of segmental histogram stretching modification which preserves the given image shape, not distorting the original shape. After global stretching, the image is partitioned into several level-sets according to threshold condition. The image information of each level-set is represented as typical value based on grouped differential values. The basic property is modified into common local schemes, thereby introducing the enhanced effect through extreme discrimination between subsets. The scheme is based on stretching the histogram of subsets in which the intensity gray levels between connected pixels are approximately same In spite of histogram widening, stretched by local image information, it neither creates nor destroys the original image, thereby preserving image shape and enhancing the contrast. By designing local histogram stretching operations, we can preserve the original shape of level-sets of the image, and also enhance the global intensity. Thus it can hold the main properties of both global and local image schemes, which leads to versatile applications in the field of digital epigraphy.

• Proceedings of the KIEE Conference
• /
• 1995.11a
• /
• pp.225-227
• /
• 1995

An object detection algorithm using a modified IDM(Image Differential Method) is proposed for detecting an object in a level crossing area. The conventional object detection method using LASER light has the deadzone that it cannot detect small objects, while the object detection method using image data in a level crossing area can detect such small objects. But the image data in a level crossing area can be changeable easily because the data is outdoor and sensitive to such surrounding environments as the change of the sun beam, the shadow of cars, and so on. So we resolve these problems by adding the normalization and the process for shadow of the image data in a level crossing area to the basic IDM(Image Differential Method).