• 한국멀티미디어학회논문지
• /
• 제25권1호
• /
• pp.1-9
• /
• 2022

Ship detection in synthetic aperture radar (SAR) images is an important application in marine monitoring for the military and civilian domains. Over the past decade, object detection has achieved significant progress with the development of convolutional neural networks (CNNs) and lot of labeled databases. However, due to difficulty in collecting and labeling SAR images, it is still a challenging task to solve SAR ship detection CNNs. To overcome the problem, some methods have employed conventional data augmentation techniques such as flipping, cropping, and affine transformation, but it is insufficient to achieve robust performance to handle a wide variety of types of ships. In this paper, we present a novel and effective approach for deep SAR ship detection, that exploits label-rich Electro-Optical (EO) images. The proposed method consists of two components: a data augmentation network and a ship detection network. First, we train the data augmentation network based on conditional generative adversarial network (cGAN), which aims to generate additional SAR images from EO images. Since it is trained using unpaired EO and SAR images, we impose the cycle-consistency loss to preserve the structural information while translating the characteristics of the images. After training the data augmentation network, we leverage the augmented dataset constituted with real and translated SAR images to train the ship detection network. The experimental results include qualitative evaluation of the translated SAR images and the comparison of detection performance of the networks, trained with non-augmented and augmented dataset, which demonstrates the effectiveness of the proposed framework.

• 한국통신학회:학술대회논문집
• /
• 한국통신학회 1987년도 춘계학술발표회 논문집
• /
• pp.35-38
• /
• 1987

The purpose of this paper is to compare a class of vector quantization techniques which include GVQ(Genera VQ) MSVQ(Mean separated VQ) and DCT_VQ The VQ techniques are applied to six images and both subjective and objective performance comparison are made The results indicate that the transform domain approach(DCT_VQ) yields more syable results than the spatial domain method (GVQ, MSVQ)

• 한국환경복원기술학회지
• /
• 제16권5호
• /
• pp.59-69
• /
• 2013

Global warming also has effects on the phenology. The limitation of phenology study is an acquisition of phenology data. Satellite images analysis can make up limitation of monitering data. This study is to analyze spatial distribution and characteristics of phenology changes using MODIS images. Research data collected images of 16 day intervals of 11 years from year 2001 to 2010. The data analyzed 228 images of 11 years. It can figure out changes of phenology by analyzing enhanced vegetation index of MODIS image. We made a comparison between changes of phenology and flowering of cherry blossoms. As a results, Startup of season spatially was getting late from southern area to north area. Startup of Phenology was foreshortened 13 days during 11 years, and change ratios of cherry blooming was getting more faster from 0.18 dat to 0.22 day per year during that same period.

• International Journal of Contents
• /
• 제13권3호
• /
• pp.1-8
• /
• 2017

Rice plant numbers and density are key factors for yield and quality of rice grains. Precise and properly estimated rice plant numbers and density can assure high yield from rice fields. The main objective of this study was to automatically detect and count rice plants using images of usual field condition from an unmanned aerial vehicle (UAV). We proposed an automatic image processing method based on morphological operation and boundaries of the connected component to count rice plant numbers after transplanting. We converted RGB images to binary images and applied adaptive median filter to remove distortion and noises. Then we applied a morphological operation to the binary image and draw boundaries to the connected component to count rice plants using those images. The result reveals the algorithm can conduct a performance of 89% by the F-measure, corresponding to a Precision of 87% and a Recall of 91%. The best fit image gives a performance of 93% by the F-measure, corresponding to a Precision of 91% and a Recall of 96%. Comparison between the numbers of rice plants detected and counted by the naked eye and the numbers of rice plants found by the proposed method provided viable and acceptable results. The $R^2$ value was approximately 0.893.

• Investigative Magnetic Resonance Imaging
• /
• 제19권1호
• /
• pp.19-30
• /
• 2015

Purpose: A new compressed sensing technique by iterative truncation of small transformed coefficients (ITSC) is proposed for fast cardiac CINE MRI. Materials and Methods: The proposed reconstruction is composed of two processes: truncation of the small transformed coefficients in the r-f domain, and restoration of the measured data in the k-t domain. The two processes are sequentially applied iteratively until the reconstructed images converge, with the assumption that the cardiac CINE images are inherently sparse in the r-f domain. A novel sampling strategy to reduce the normalized mean square error of the reconstructed images is proposed. Results: The technique shows the least normalized mean square error among the four methods under comparison (zero filling, view sharing, k-t FOCUSS, and ITSC). Application of ITSC for multi-slice cardiac CINE imaging was tested with the number of slices of 2 to 8 in a single breath-hold, to demonstrate the clinical usefulness of the technique. Conclusion: Reconstructed images with the compression factors of 3-4 appear very close to the images without compression. Furthermore the proposed algorithm is computationally efficient and is stable without using matrix inversion during the reconstruction.

• 대한임베디드공학회논문지
• /
• 제16권5호
• /
• pp.225-232
• /
• 2021

Deraining is one of the image restoration tasks and should consider a tradeoff between local details and broad contextual information while recovering images. Current studies adopt an attention mechanism which has been actively researched in natural language processing to deal with both global and local features. This paper classifies existing deraining methods and provides comparative analysis and performance comparison by using several datasets in terms of generalization.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2000년도 춘계 학술대회 논문집 통권 3호 Proceedings of the 2000 KSRS Spring Meeting
• /
• pp.49-54
• /
• 2000

Conventional methods for measuring winds provide wind velocity observations over limited area and time period. The use of satellite imagery for measuring wind velocity overcomes some of these limitations by providing wide area and near condinuous coverage. And its accurate depiction is essential for operational weather forecasting and for initialization of NWP models. GMS-5 provides full disk images at hourly intervals. At four times each day - 0500, 1100, 1700, 2300 hours UTC-a series of three images is received, separated by thirty minutes, centered at the four times. The current wind system generates winds from sets of 3 infrared(IR) images, separated by an hour, four times a day. It also produces visible(VIS) and water vapor(WV) image-based winds from half-hourly imagery four times a day. The derivation of wind from satellite imagery involves the identification of suitable cloud targets. tracking the targets on sequential images, associating a pressure height with the derived wind vector, and quality control. The aim of this research is to incorporate imagery from other available spectral channels and examine the error characteristics of winds derived from these images.

• Imaging Science in Dentistry
• /
• 제40권2호
• /
• pp.59-62
• /
• 2010

Purpose : To investigate and compare the prevalence, size, and location of maxillary sinus septa on panoramic and cone beam computed tomography (CBCT) images. Materials and Methods : Two hundred patients who had taken both panoramic and CBCT images were included. The location of maxillary sinus septa on the panoramic radiographs were recorded and confirmed on the CBCT images. Also the size of septa was measured on the reformatted CBCT images. Results : The prevalence of the patients who had maxillary sinus septa was 51.0% and they showed 179 septa totally. Among them 51.0% of the patients had one septum, 32.4% two septa, 13.7% three, and 2.9% four. The measured heights of the septa were $4.37{\pm}2.87\;mm$, $3.51{\pm}2.47\;mm$, and $3.04{\pm}2.37\;mm$ in the medial, middle, and lateral areas, respectively. It was revealed that 1.0% was located at canine region, 18.0% at first premolar, 25.0% at second premolar, 22.7% at first molar, 19.8% at second molar, and 14.0% at third molar region. Among 213 septa depicted by the panoramic radiographs, only 69.0% were confirmed at the CBCT images. Conclusion : Since various heights and courses of the septa can develop in all parts of the maxillary sinus, adequate assessment of the inner aspect of the maxillary sinus is essential to avoid complications during sinus augmentation procedures. CBCT scanning is the preferred radiographic method for detecting the presence of sinus septa.

• IEIE Transactions on Smart Processing and Computing
• /
• 제1권1호
• /
• pp.17-26
• /
• 2012

This paper presents an image processing-based validation method for unrecognizable numbers in severely distorted license plate images which have been degraded by various factors including low-resolution, low light-level, geometric distortion, and periodic noise. Existing vehicle license plate recognition (LPR) methods assume that most of the image degradation factors have been removed before performing the recognition of printed numbers and letters. If this is not the case, conventional LPR becomes impossible. The proposed method adopts a novel approach where a set of reference number images are intentionally degraded using the same factors estimated from the input image. After a series of image processing steps, including geometric transformation, super-resolution, and filtering, a comparison using cross-correlation between the intentionally degraded reference and the input images can provide a successful identification of the visually unrecognizable numbers. The proposed method makes it possible to validate numbers in a license plate image taken under low light-level conditions. In the experiment, using an extended set of test images that are unrecognizable to human vision, the proposed method provides a successful recognition rate of over 95%, whereas most existing LPR methods fail due to the severe distortion.

• 한국의류학회지
• /
• 제46권2호
• /
• pp.321-330
• /
• 2022

This study analyzed the difference between 2D image and 3D virtual clothing images based on stripe arrangement to obtain fundamental data for slim appearance. First, the slimming effect according to the three types of stripe ratio was examined. Subsequently, the slimming effects of seven types of one-piece dress designs according to the stripe location were analyzed. Subjective ranking was evaluated. The width items and radius of curvature were measured for the image's respective parts. Consequently, in 2D image and 3D virtual clothing images, the one with the narrowest stripe ratio was evaluated as the slimmest; however, the conditions for the slimming effect were different. In the seven one-piece dress designs, a difference was apparent in the ranking of the 2D image and 3D virtual clothing images. In the 3D virtual clothing image, arranging the stripes on the entire garment proved inefficient. The stripes were curved according to the curvature of the human body, creating an optical illusion that differed from that of the 2D image.