• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.100-102
• /
• 2003

This system is capable of obtaining quantitative information from images using natural features on the ortho-image maps that correspond with those from topographical maps. However, the qualitative information can also be obtained because because of the excellent visibility of ortho-image maps. There are plenty of promise for the use of ortho-image maps in the next generation topographic technology because of its wider applicability within the field. In keeping with the cutting edge, we produced ortho-image maps by scanning a specified area in narrow sections using the PKNU 2: a multispectral digital aerial photographing system made by ourselves. We evaluated the precision of the ortho-image maps, and performed an evaluation of the PKNU 2 system's capacity to improve the equipment of the PKNU 2. Ortho-image maps were made using Ground Control Points (GCPs) which were obtained from digital maps and aerial photographs of the PKNU 2. Thus, we demonstrated that it was possible to produce the ortho-image maps, which has a good constant level rate of less than 1m. The PKNU 2 system needs to be improving in the sensitivity of level maintenance equipment in the evaluation in terms of performance. It is thus required to survey the GCPs precisely for an accurate study.

• Journal of applied mathematics & informatics
• /
• v.9 no.2
• /
• pp.793-806
• /
• 2002

Feature-based similarity retrieval become an important research issue in image database systems. The features of image data are useful to discrimination of images. In this paper, we propose the highspeed k-Nearest Neighbor search algorithm based on Self-Organizing Maps. Self-Organizing Maps (SOM) provides a mapping from high dimensional feature vectors onto a two-dimensional space. The mapping preserves the topology of the feature vectors. The map is called topological feature map. A topological feature map preserves the mutual relations (similarity) in feature spaces of input data. and clusters mutually similar feature vectors in a neighboring nodes. Each node of the topological feature map holds a node vector and similar images that is closest to each node vector. In topological feature map, there are empty nodes in which no image is classified. We experiment on the performance of our algorithm using color feature vectors extracted from images. Promising results have been obtained in experiments.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.19 no.5
• /
• pp.699-712
• /
• 1995

The purpose of this study was 1) to identify the constructing factors of jeans brand image 2) to analyze the domestic jeans market using perceptual maps of three benefit segments based on stdy(I). The questionnaire consisted of brand preference, attribute of brand image and wearer image was selected from the previous studies or developed for this study. The subjects were 350 male and female university students who have purchased at least one of the nine jeans wear brand selected for the study. For statistical analysis, reliability test, factor analysis, MANOVA, and multiple regression were used. The results of this study were as follows: 1. Symbolism, quality, and economy were found out as constricting factors of brand image in the attribute dimensions, while innovative and active image were found out in the wearer image dimensions. 2. 9 Perceptual maps of attribute dimensions and 3 perceptual maps of wearer image dimensions were constructed and each ideal vector was drawn.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.1
• /
• pp.133-138
• /
• 2009

We investigate the properties of meet preserving maps on strictly two-sided, commutative biquantales. Using the properties of meet preserving maps, we induce the Zadeh image and pre image operators.

• Journal of the Korean Home Economics Association
• /
• v.29 no.4
• /
• pp.45-64
• /
• 1991

Each of us has personal and unique "Mental Images" of environment that are the results of a two-way process between an observer and his environment. By understanding of people's image on their physical settings through cognitive maps, congruent with their design environment. In this context, this study is to fine out the characteristics of the resident's image in apartment housing that are modern housing. This paper is theoretically based on Appleyard's cognitive maps types, Lynch's the components of image, Harrison and Howard's image enhancing factors and Applyard's image enhancing factors of buildings. For the purpose of this study, two-instruments were used. One was sketch mapping, that told of visual aspects of cognitive map. Another was verbal questionnaire, which was composed of nonvisual aspects of image components and image enhancing factors. For the methods of research analysis, Percentage, Frequency, and Chi-square test were used. The results of this study are as follows. Firstly, as for the types of cognitive maps, the rate of sequential maps and spatial maps are nearly same. Secondly, as for the components of images, landmarks and districts are significant elements in apartment housing, and edges, nodes and paths are rare element. Thirdly, with regard to image enhancing factors, landscaping is shown to be most effective at landmarks and districts. And finally, in examing the differences of images between various social groups, significant variables are years of living and housing size. Age and income are the next. In this study of image, we can learn about manner in which individuals come to see, understand, and cope with the environment at a personal level. So, user-oriented approach is useful in environmental design, and particularly, it is useful in the apartment housing of the future life space.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2004.11a
• /
• pp.239-243
• /
• 2004

Recently various thematic maps and image maps using aerial photograph and satellite imagery are frequently made. The geo-referencing is essential to make image map and topographic map using aerial photograph and satellite imagery. For this geo-referencing, Ground Control Points (GCPs) are needed. In this paper, we used GPS relative positioning to measure GCP ground coordinate and the accuracy of 8cm level was achieved. We made GCP image chips for the efficiency of geo-referencing and carried out the bundle adjustment of aerial photographs using GCP image chips to acquire the GCP photo coordinate with image matching technique. Finally we analyzed the accuracy of bundle adjustment compared to the accuracy of the case in using digital maps to acquire GCP photo coordinate.

• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.238-242
• /
• 2007

Today's commercial high resolution satellite imagery such as that provided by IKONOS and QuickBird, offers the potential to extract useful spatial information for geographical database construction and GIS applications. Digital maps supply the most generally used GIS data probiding topography, road, and building information. Currently, the building information provided by digital maps is incompletely constructed for GIS applications due to planar position error and warped shape. We focus on extracting of the accurate building information including position, shape, and height to update the building information of the digital maps and GIS database. In this paper, we propose a new method of 3D building information extraction with a single high resolution satellite image and digital map. Co-registration between the QuickBird image and the 1:1,000 digital maps was carried out automatically using the RPC adjustment model and the building layer of the digital map was projected onto the image. The building roof boundaries were detected using the building layer from the digital map based on the satellite azimuth. The building shape could be modified using a snake algorithm. Then we measured the building height and traced the building bottom automatically using triangular vector structure (TVS) hypothesis. In order to evaluate the proposed method, we estimated accuracy of the extracted building information using LiDAR DSM.

• Smart Media Journal
• /
• v.12 no.9
• /
• pp.45-59
• /
• 2023

Weakly supervised object localization (WSOL) is a task of localizing an object in an image using only image-level labels. Previous studies have followed the conventional class activation mapping (CAM) pipeline. However, we reveal the current CAM approach suffers from problems which cause original CAM could not capture the complete defects features. This work utilizes a convolutional neural network (CNN) pretrained on image-level labels to generate class activation maps in a multi-scale manner to highlight discriminative regions. Additionally, a vision transformer (ViT) pretrained was treated to produce multi-head attention maps as an auxiliary detector. By integrating the CNN-based CAMs and attention maps, our approach localizes defective regions without requiring bounding box or pixel-level supervision during training. We evaluate our approach on a dataset of apple images with only image-level labels of defect categories. Experiments demonstrate our proposed method aligns with several Object Detection models performance, hold a promise for improving localization.

• Journal of Korea Society of Industrial Information Systems
• /
• v.7 no.1
• /
• pp.81-88
• /
• 2002

This paper proposes an extraction method which extracts two different thematic maps, which have different line thickness from each other in a raster image that contains the two thematic maps. In the proposed method, the depth of each pixel is calculated according to the amount of pixels in its surrounding neighborhood, and then the thinning is performed. By using depth threshold, two thematic maps are first extracted from the thinning result. There are noise images and skeleton disconnection in the lines of each extracted thematic map. Each thematic map extraction is finally completed after removing the noise images and connecting the disconnected lines. Through the experiment, we showed that the proposed method could be used for the extraction of each thematic map of a raster image which included two thematic maps.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.3942-3961
• /
• 2019

Recently, the models of deep super-resolution networks can successfully learn the non-linear mapping from the low-resolution inputs to high-resolution outputs. However, for large scaling factors, this approach has difficulties in learning the relation of low-resolution to high-resolution images, which lead to the poor restoration. In this paper, we propose Stage Generative Adversarial Networks (Stage-GAN) with semantic maps for image super-resolution (SR) in large scaling factors. We decompose the task of image super-resolution into a novel semantic map based reconstruction and refinement process. In the initial stage, the semantic maps based on the given low-resolution images can be generated by Stage-0 GAN. In the next stage, the generated semantic maps from Stage-0 and corresponding low-resolution images can be used to yield high-resolution images by Stage-1 GAN. In order to remove the reconstruction artifacts and blurs for high-resolution images, Stage-2 GAN based post-processing module is proposed in the last stage, which can reconstruct high-resolution images with photo-realistic details. Extensive experiments and comparisons with other SR methods demonstrate that our proposed method can restore photo-realistic images with visual improvements. For scale factor ${\times}8$, our method performs favorably against other methods in terms of gradients similarity.