• 한국측량학회지
• /
• 제38권1호
• /
• pp.23-33
• /
• 2020

최근 UAV (Unmanned Aerial Vehicle)를 이용하여 고해상도 영상을 편리하게 취득할 수 있게 되면서 저비용으로 소규모 지역의 관측 및 공간정보 제작이 가능하게 되었다. 특히, 농업환경 모니터링을 위하여 작물생산 지역의 피복지도 생성에 대한 연구가 활발히 진행되고 있으며, 랜덤 포레스트와 SVM (Support Vector Machine) 및 CNN(Convolutional Neural Network) 을 적용하여 분류 성능을 비교한 결과 영상분류에서 딥러닝 적용에 대하여 활용도가 높은 것으로 나타났다. 특히, 위성영상을 이용한 피복분류는 위성영상 데이터 셋과 선행 파라메터를 사용하여 피복분류의 정확도와 시간에 대한 장점을 가지고 있다. 하지만, 무인항공기 영상은 위성영상과 공간해상도와 같은 특성이 달라 이를 적용하기에는 어려움이 있다. 이러한 문제점을 해결하기 위하여 위성영상 데이터 셋이 아닌 UAV를 이용한 데이터 셋과 국내의 소규모 복합 피복이 존재하는 농경지 분석에 활용이 가능한 딥러닝 알고리즘 적용 연구를 수행하였다. 본 연구에서는 최신 딥러닝의 의미론적 영상분류인 DeepLab V3+, FC-DenseNet (Fully Convolutional DenseNets), FRRN-B (Full-Resolution Residual Networks) 를 UAV 데이터 셋에 적용하여 영상분류를 수행하였다. 분류 결과 DeepLab V3+와 FC-DenseNet의 적용 결과가 기존 감독분류보다 높은 전체 정확도 97%, Kappa 계수 0.92로 소규모 지역의 UAV 영상을 활용한 피복분류의 적용가능성을 보여주었다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2005년도 추계종합학술대회
• /
• pp.561-564
• /
• 2005

In this paper, we present a new JPEG (Joint Photograph Experts Group) compression architecture which compresses still image into fixed size of bitstream to use restricted system memory efficiently. The size of bitstream is determined by the complexity of image and the quantization table. But the quantization table is set in advance the complexity of image is the essential factor. Therefore the size of bitstream for high complexity image is large and the size for low complexity image is small. This means that the management of restricted system memory is difficult. The proposed JPEG encoder estimates the size of bitstream using the correlation between consecutive frames and selects the quantization table suited to the complexity of image. This makes efficient use of system memory.

• 패션비즈니스
• /
• 제19권6호
• /
• pp.14-27
• /
• 2015

This study explores the comparison of body image, body satisfaction, and clothing behaviors between Korean and British young women. Body image was measured by two methods: visual and verbal. For the data analysis, the Statistical Package for Social Science (SPSS) Version 16.0 for Windows was used to provide descriptive statistics, an independent sample t-test, and paired sample t- tests were applied in this study. Our results show that Korean and British female college students perceived ideal-body images that were smaller than their self defined body images. The ideal and self-images were significantly different in both groups. Both groups were dissatisfied with their own body size. The study was limited to a small sample size. Future studies using more participants from a more diverse age group and ethnic groups are recommended. The study will help marketers and retailers develop new products and new markets aimed at Korean and British women related to body image and body satisfaction.

• 한국ITS학회 논문지
• /
• 제8권1호
• /
• pp.127-135
• /
• 2009

This paper presents a new technique for corner shape based object retrieval from a database. The proposed feature matrix consists of values obtained through a neighborhood operation of detected corners. This results in a significant small size feature matrix compared to the algorithms using color features and thus is computationally very efficient. The corners have been extracted by finding the intersections of the detected lines found using Hough transform. As the affine transformations preserve the co-linearity of points on a line and their intersection properties, the resulting corner features for image retrieval are robust to affine transformations. Furthermore, the corner features are invariant to noise. It is considered that the proposed algorithm will produce good results in combination with other algorithms in a way of incremental verification for similarity.

• 한국멀티미디어학회논문지
• /
• 제19권7호
• /
• pp.1116-1126
• /
• 2016

Image segmentation is widely used in the pre-processing stage of image analysis and, therefore, the accuracy of image segmentation is important for performance of an image-based analysis system. An efficient image segmentation method is proposed, including a filtering process for super-pixels, improved saliency map information, and a merge process. The proposed algorithm removes areas that are not equal or of small size based on comparison of the area of smoothed superpixels in order to maintain generation of a similar size super pixel area. In addition, application of a bilateral filter to an existing saliency map that represents human visual attention allows improvement of separation between objects and background. Finally, a segmented result is obtained based on the suggested merging process without any prior knowledge or information. Performance of the proposed algorithm is verified experimentally.

• 정보저장시스템학회논문집
• /
• 제7권2호
• /
• pp.75-79
• /
• 2011

Recently mobile phone camera become generally spread, it is required to develop high resolution, multi-functional camera module for obtaining high image quality. To satisfy this demand, number of pixels has been increased and pixel size decreased in small mobile phone cameras. As a result, image quality is seriously dropped by blur phenomena. Especially when hand tremble is occurred, image quality is dropped by camera shake. Therefore, to obtain high quality image, it is necessary to compensate user's hand tremble. In this paper, we propose voice coil actuator for compensating hand tremble, which can apply optical image stabilization (OIS) system. Sensitivity analysis and size optimization are performed to obtain high driving force. Finally, it is confirmed that the optimized electromagnetic circuit can be applied in OIS system.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2007년도 7th International Meeting on Information Display 제7권1호
• /
• pp.617-620
• /
• 2007

The temporal dark image sticking phenomena for both the face-to-face and coplanar sustain electrode structures were compared. For both structures, the temporal dark image sticking phenomena were examined by measuring the difference in the IR emission, display luminance, perceived luminance, and temperature between the image sticking and the no image sticking cells. For the face-to-face structure, the 10-min sustain discharge causes a small increment of the panel temperature thanks to the ITO-less electrode structure, thereby resulting in mitigating the temporal dark image sticking phenomenon.

• 반도체디스플레이기술학회지
• /
• 제9권4호
• /
• pp.51-56
• /
• 2010

As the price of image acquisition component becomes low enough, the compact and easily accessible handheld devices are generally equipped with image acquisition functionality. This trend speeds up various applications in diverse areas such as image related services and software. Therefore users strongly need to identify their images effectively and efficiently so that the duplicated images are perceived as one physical entity. In order to handle this environment, we propose a number of methods that generate image identifiers utilizing fundamental image features. In this paper, we analyze the identifier generation methods in terms of various size patterns, especially for tiny size cases, since the small images does not contain abundant pixels for feature extraction. In this paper, experimental evaluation over identifier generation methods' behavior according to different sizes is demonstrated.

• 센서학회지
• /
• 제30권5호
• /
• pp.286-289
• /
• 2021

Image classification is one of the fundamental applications of computer vision. It enables a system to identify an object in an image. Recently, image classification applications have broadened their scope from computer applications to edge devices. The convolutional neural network (CNN) is the main class of deep learning neural networks that are widely used in computer tasks, and it delivers high accuracy. However, CNN algorithms use a large number of parameters and incur high computational costs, which hinder their implementation in edge hardware devices. To address this issue, this paper proposes a lightweight image classifier that provides good accuracy while using fewer parameters. The proposed image classifier diverts the input into three paths and utilizes different scales of receptive fields to extract more feature maps while using fewer parameters at the time of training. This results in the development of a model of small size. This model is tested on the CIFAR-10 dataset and achieves an accuracy of 90% using .26M parameters. This is better than the state-of-the-art models, and it can be implemented on edge devices.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.2059-2064
• /
• 2004

Texture analysis is an important role for automatic visual insfection. This paper presents an application of wavelet extension and Gray level co-occurrence matrix (GLCM) for detection of defect encountered in textured images. Texture characteristic in low quality images is not to easy task to perform caused by noise, low frequency and small dimension. In order to solve this problem, we have developed a procedure called wavelet image extension. Wavelet extension procedure is used to determine the frequency bands carrying the most information about the texture by decomposing images into multiple frequency bands and to form an image approximation with higher resolution. Thus, wavelet extension procedure offers the ability to robust feature extraction in images. Then the features are extracted from the co-occurrence matrices computed from the sub-bands which performed by partitioning the texture image into sub-window. In the detection part, Mahalanobis distance classifier is used to decide whether the test image is defective or non defective.