• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.9
• /
• pp.2458-2482
• /
• 2023

Medical image segmentation techniques based on convolution neural networks indulge in feature extraction triggering redundancy of parameters and unsatisfactory target localization, which outcomes in less accurate segmentation results to assist doctors in diagnosis. In this paper, we propose a multi-level semantic-rich encoding-decoding network, which consists of a Pooling-Conv-Former (PCFormer) module and a Cbam-Dilated-Transformer (CDT) module. In the PCFormer module, it is used to tackle the issue of parameter explosion in the conservative transformer and to compensate for the feature loss in the down-sampling process. In the CDT module, the Cbam attention module is adopted to highlight the feature regions by blending the intersection of attention mechanisms implicitly, and the Dilated convolution-Concat (DCC) module is designed as a parallel concatenation of multiple atrous convolution blocks to display the expanded perceptual field explicitly. In addition, MultiHead Attention-DwConv-Transformer (MDTransformer) module is utilized to evidently distinguish the target region from the background region. Extensive experiments on medical image segmentation from Glas, SIIM-ACR, ISIC and LGG demonstrated that our proposed network outperforms existing advanced methods in terms of both objective evaluation and subjective visual performance.

• Journal of Korea Multimedia Society
• /
• v.14 no.12
• /
• pp.1528-1535
• /
• 2011

We present a fog generation algorithm in 2D images. The proposed algorithm provides a scattering model for the approximated calculation of fog density. The scattering model needs parameters of distance and altitude information. However, 2D images do not include that information, so that we calculate them from the depth information generated in an interactive manner, and estimate the scattering factor by using the scattering model. Then we generate fog effect on an input image using the scattering factor by distance-oriented selection blur and color blending. With the algorithm, we can easily create the fog-effected images and fog generated animation from 2D images.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.28 no.910
• /
• pp.1231-1241
• /
• 2004

The purpose of this study was to investigate the musicians' body image and fashion styles portrayed in Korea and American music' video in aspects of body dimension such as sex, race, nature, and technology. The results were as follows. In aspects of sex, male musicians showed the blurring of gender appearance, while female musicians reflected sexuality which intensified the gender norms. In aspects of race, musicians pursued the racial blending appearances. In nature and technology dimension, Korea music video focused on natural imagery, while American music video showed a lot of technology imagies. And, the body combined with the nature and technology implied the post human image in the future.

• Journal of Information Processing Systems
• /
• v.9 no.4
• /
• pp.660-677
• /
• 2013

The blending of defective texture with the ambience texture results in camouflage. The gray value or color distribution pattern of the camouflaged images fails to reflect considerable deviations between the camouflaged object and the sublimating background demands improved strategies for texture analysis. In this research, we propose the implementation of an initial enhancement of the image that employs line masks, which could result in a better discrimination of the camouflaged portion. Finally, the gray value distribution patterns are analyzed in the enhanced image, to fix the camouflaged portions.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.1468-1471
• /
• 2012

It is always a dream to recreate the experience of a particular place, the Panorama Virtual Reality has been interpreted as a kind of technology to create virtual environments and the ability to maneuver angle for and select the path of view in a dynamic scene. In this paper we examined an efficient algorithm for Image registration and stitching of captured imaged from a video stream. Two approaches are studied in this paper. First, dynamic programming is used to spot the ideal key points, match these points to merge adjacent images together, later image blending is use for smooth color transitions. In second approach, FAST and SURF detection are used to find distinct features in the images and a nearest neighbor algorithm is used to match corresponding features, estimate homography with matched key points using RANSAC. The paper also covers the automatically choosing (recognizing, comparing) images to stitching method.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.04a
• /
• pp.352-354
• /
• 2012

본 논문에서는 웹캠 카메라 어레이(camera array)로 얻은 여러 장의 이미지를 빠른 속도로 봉합(stitching)하여 고해상도 이미지를 얻기 위해 그래픽스 하드웨어를 이용하는 병렬 알고리즘을 제시한다. 고정된 레이아웃의 카메라 어레이를 이용하여 평면 혹은 원경을 촬영하는 경우, 기존에 널리 쓰이던 평면 사영 이미지 봉합(planar projective image stitching)과 선형 혼합(linear blending)을 통해 만족스런 결과를 얻을 수 있다. 본 논문에서는 이러한 연산을 그래픽스 하드웨어에서 병렬처리 함으로써 추후 실시간 고해상도 동영상 스트리밍 이미지 병합에 활용할 수 있을 정도로 빠른 속도로 처리하는 방법을 제시한다.

• Fire Science and Engineering
• /
• v.30 no.6
• /
• pp.92-98
• /
• 2016

This study evaluated thermal image real-time estimation and fire alarm using by a CCD camera, which has been a seamless feature-point analysis method, according to the angle and position and image fusion by a vector coordinate point set-up of equal shape. The system has higher accuracy, fixing data value of temperature sensing and fire image of 0~255, and sensor output-value of 0~5,000. The operation time of a flame specimen within 500 m, 1000 m, and 1500 m from the test report specimen took 7 s, 26 s, and 62 s, respectively, and image creation was proven. A diagnosis of fire accident was designated to 3 steps: Caution/Alarm/Fire. Therefore, a series of process and the transmission of SNS were identified. A light bulb and fluorescent bulb were also tested for a false alarm test, but no false alarm occurred. The possibility that an unwanted alarm will be reduced was verified through a forecast of the fire progress or real-time estimation of a thermal image by the change in the image of a time-based flame and an analysis of the diffusion velocity.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.9
• /
• pp.1738-1743
• /
• 2017

This paper presents an efficient image stitching method using preprocessing in order to generate a super resolution image. Two-dimensional (2D) scanners are consistently used in various areas but they have limitations such as paper sizes and materials. To overcome these problem with low-cost, an efficient imaging stitching method is proposed for producing a super resolution panorama image. To scan a very large sized paper using mobile phones, a simple portable cradle which fixes height is employed producing an input image set. To improve matching performance, a preprocessing method is introduced before searching correspondences. Then alpha blending is applied to an input image set to produce a super resolution panorama image. The proposed method is faster and easier than the existing method which is employed by Open CV. Experiment results show that the proposed method is three times faster and performs better than the existing method.

• The Research Journal of the Costume Culture
• /
• v.15 no.2 s.67
• /
• pp.221-231
• /
• 2007

In the modern society of the 21st century where the visuality is emphasized, the illustration which expresses one symbol by embodying it is getting more spotlight than the use of character. Specifically, the fashion illustration serves the role of image promotion in other fields, for such as book, magazine, etc, as well as the fashion, and is taking root in daily lives. In response to that, this study researched on the fusion style fashion and the fashion illustration by using the violet color. The violet color which symbolizes the mysticism is being applied to society and culture in various forms, and has been considered as a precious color that represents god since ancient time, and still possesses its mysterious image even in the modern times. Moreover, as the appearance of fusion styled clothes which consolidates the latest trend image of the west and east is combined with the mysterious image of the purple to express new cultural image beyond highlighting specific image element of a country, the study on the fashion illustration based on the clothes based on the purple and fusion style would be very interesting. This study tried to produce the work that aims at the practical study based on the purple & fusion style and the theoretical consideration of fashion illustration through the analysis and investigative study of preceding studies. The findings from the study is like this: First, the violet the color, which represents the mysticism, can be confirmed to be an sophisticated, elegant and attractive image color that leads the modern fashion. Second, the fusion style which is coming into the spotlight recently can be found to have a great visual effect of consolidating the image of the west and east through the mixture with the purple color. Third, it was confirmed in our production of works that the fashion illustration was expressing a mysterious and new fusion style by blending appropriately with the purple color.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.3
• /
• pp.439-445
• /
• 2021

Low-cost, ultra-high-speed cameras, made possible by the development of image sensors and small displays, can be very useful in image processing and pattern recognition. This paper introduces an algorithm that corrects irregular lighting from a high-speed image that is continuously input with a slight time interval, and which then obtains an exposed skin color region that is the area of interest in a person from the corrected image. In this study, the non-uniform lighting effect from a received high-speed image is first corrected using a frame blending technique. Then, the region of interest is robustly obtained from the input high-speed color image by applying an elliptical skin color distribution model generated from iterative learning in advance. Experimental results show that the approach presented in this paper corrects illumination in various types of color images, and then accurately acquires the region of interest. The algorithm proposed in this study is expected to be useful in various types of practical applications related to image recognition, such as face recognition and tracking, lighting correction, and video indexing and retrieval.