• 대한원격탐사학회지
• /
• 제24권3호
• /
• pp.257-266
• /
• 2008

This study presents an approach to synthesize multispectral images at a higher resolution by exploiting a high-resolution image acquired in panchromatic modality. The synthesized images should be similar to the multispectral images that would have been observed by the corresponding sensor at the same high resolution. The proposed scheme is designed to reconstruct the multispectral images at the higher resolution with as less color distortion as possible. It uses a regression model of the second order to fit panchromatic data to multispectral observations. Based on the regression model, the multispectral images at the higher spatial resolution of the panchromatic image are optimized by a quadratic programming. In this study, the new method was applied to the IKONOS 1m panchromatic and 4m multispectral data, and the results were compared with them of several current approaches. Experimental results demonstrate that the proposed scheme can achieve significant improvement over other methods.

• 한국측량학회지
• /
• 제36권5호
• /
• pp.413-422
• /
• 2018

Many applications using satellite data from high-resolution multispectral sensors require an image fusion step, known as pansharpening, before processing and analyzing the multispectral images when spatial fidelity is crucial. Image fusion methods are to improve images with higher spatial and spectral resolutions by reducing spectral distortion, which occurs on image fusion processing. The image fusion methods can be classified into MRA (Multi-Resolution Analysis) and CSA (Component Substitution Analysis) approaches. To suggest the efficient image fusion method for Pleiades and KOMPSAT (Korea Multi-Purpose Satellite) 3 satellites, this study will evaluate image fusion methods for multispectral and panchromatic images. HPF (High-Pass Filtering), SFIM (Smoothing Filter-based Intensity Modulation), GS (Gram Schmidt), and GSA (Adoptive GS) were selected for MRA and CSA based image fusion methods and applied on multispectral and panchromatic images. Their performances were evaluated using visual and quality index analysis. HPF and SFIM fusion results presented low performance of spatial details. GS and GSA fusion results had enhanced spatial information closer to panchromatic images, but GS produced more spectral distortions on urban structures. This study presented that GSA was effective to improve spatial resolution of multispectral images from Pleiades 1A and KOMPSAT 3.

• 대한원격탐사학회지
• /
• 제24권5호
• /
• pp.525-533
• /
• 2008

This study presents a pansharpening method, so called FitPAN, to synthesize multispectral images at a higher resolution by exploiting a high-resolution image acquired in panchromatic modality. FitPAN is a modified version of the quadratic programming approach proposed in (Lee, 2008), which is designed to generate synthesized multispectral images similar to the multispectral images that would have been observed by the corresponding sensor at the same high resolution. The proposed scheme aims at reconstructing the multispectral images at the higher resolution with as less spectral distortion as possible. This study also proposes a sharpening process to eliminate some distortions appeared in the fused image of the higher resolution. It employs the Point-Jacobian MAP iteration utilizing the contextual information of the original panchromatic image. In this study, the new method was applied to the IKONOS 1m panchromatic and 4m multispectral data, and the results were compared with them of several current approaches. Experimental results demonstrate that the proposed scheme can achieve significant improvement in both spectral and block distortion.

• 한국항해학회지
• /
• 제24권4호
• /
• pp.227-234
• /
• 2000

In general, neural networks are widely used for the category classification of multispectral images. Since the input multispectral images into neural networks we, however, low contrast images, neural networks converge very slowly and are of bad performance. To overcome this problem, we propose a new image enhancement method which consists of smoothing process, finding the main valley and enhancement process. In addition the enhanced images by the proposed method are used as the input of neural networks for the category classification. When the new category classification method is applied to multispectral LANDSAT TM images, we verified that the neural networks converge very lastly and that the overall category classification performance is improved.

• 대한원격탐사학회지
• /
• 제22권2호
• /
• pp.101-110
• /
• 2006

본 연구에서는 항공 관측으로 얻어진 다중분광영상과 LIDAR (LIght Detection And Ranging) 자료를 이용하여 농업지역의 토지피복 분류 정도를 분석하였다. 다중분광영상은 녹색, 적색, 근적외역의 3분광으로 이루어져 있다. LIDAR 벡터 자료로부터 최초 반사강도 영상과 최초 반사 표고 자료와 최후 반사의 지상 표고 자료의 차이로 산출된 식생 높이 영상이 얻어졌다. 토지피복 분류 방법은 최대우도법을 사용했으며, 다중분광영상의 3밴드 영상 LIDAR의 반사강도 영상, 식생 높이 영상을 이용하였다. 모든 영상을 이용한 토지피복 분류의 전체 정도는 85.6%로 다중분광영상만을 이용한 정도보다 10%이상 향상되었다. 여러 농작물간의 높이의 차이, 수목과 농작물 높이의 차이와 LIDAR 반사강도 차이로 인하여 다중분광영상과 LIDAR 영상을 사용한 토지피복 분류의 정도가 향상되었다.

• 전자공학회논문지B
• /
• 제33B권8호
• /
• pp.42-49
• /
• 1996

In this paper, we propose a satellite multispectral image data compression method using classified vector quantization. This method classifies each pixel vector considering band characteristics of multispectral images. For each class, we perform both intraband and interband vector quantization to romove spatial and spectral redundancy, respectively. And residual vector quantization for error images is performed to reduce error of interband vector quantization. Thus, this method improves compression efficiency because of removing both intraband(spatial) and interband (spectral) redundancy in multispectral images, effectively. Experiments on landsat TM multispectral image show that compression efficiency of proposed method is better than that of conventional method.

• 한국지리정보학회지
• /
• 제9권1호
• /
• pp.46-55
• /
• 2006

본 연구에서는 항공기로부터 얻어지는 다수의 다중 분광영상을 자동적인 모자이킹 방법을 개발함으로써 수작업을 최대한 줄이는데 목적을 두었다. DuncanTech MS4100 카메라를 이용하여 2436개의 녹색, 적색, 근적외 삼분광 영상이 획득되었다. 카메라 영상과 함께 관측한 LIDAR(LIght Detection And Ranging)자료와 항공기의 위치와 자세를 측정하기위해 GPS/INS(global positioning system/inertial navigation system)자료도 산출되었다. 다수의 다중 분광 영상은 우선 무감독 분류를 적용하여 영상 패턴으로 변환하였다. 인접한 영상의 패턴을 비교하여 각 영상의 상대적인 공간의 위치를 파악하였다. 모든 항공 영상 중에서 80%의 인접한 영상 패턴의 일치율을 파악하고 모자이킹할 수 있었다. 다음으로 GPS/INS자료와 무감독 분류를 혼합한 방법으로 항공 영상을 자동 모자이킹 수행하였다. GPS/INS자료와 영상 포착시점의 불일치로 연속되는 GPS/INS자료 중에 무감독 분류를 이용한 영상 패턴의 일치율을 조사하여 영상포착시점에 일치하는 GPS/INS자료를 선택하였다. 이 혼합방법으로 96%의 영상을 모자이킹했으며, LIDAR자료와의 검정에서 공간적 정도 RMSE는 1.44 m에 불과했다.

• 조명전기설비학회논문지
• /
• 제30권2호
• /
• pp.1-10
• /
• 2016

A system which provides multispectral near-IR and visible gray images of objects is constructed and an algorithm is derived to acquire a natural color image of objects from the gray images. A color image of 24 color patches is obtained by recovering their CIE (International Commission on Illumination) LAB color coordinates $L^*$, $a^*$, $b^*$ from their gray images using the algorithm based on polynomial regression. The system is composed of a custom-designed LED illuminator emitting multispectral near-IR illuminations, fluorescent lamps and a monochrome digital camera. Color reproducibility of the algorithm is estimated in CIELAB color difference ${\Delta}E^*_{ab}$. And as a result, if yellow and magenta color patches with around 10 ${\Delta}E^*_{ab}$ are disregarded, the average ${\Delta}E^*_{ab}$ is 2.9, and this value is within the acceptability tolerance for quality evaluation for digital color complex image.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2007년도 춘계학술발표회 논문집
• /
• pp.207-210
• /
• 2007

본 연구에서는 고해상도의 IKONOS panchromatic 영상과 multispectral 영상을 IHS와 PCA 방법으로 중합하고 그 결과를 비교하였다. 평가에 있어서는 중합된 영상들과 원영상간의 필셀 값에 대한 평균제곱근오차를 구하고 그 결과를 분석하였다. 분석 결과, multispectral band 1, 3, 4를 사용하는 IHS 방법, multispectral band 1, 2, 4를 사용하는 IHS 방법 및 multispectral band 1, 3, 4를 사용하는 PCA 방법이 원영상의 특성을 잘 보존하는 것으로 평가되었다.

• Journal of Biosystems Engineering
• /
• 제38권3호
• /
• pp.199-207
• /
• 2013

Purpose: Tomatoes, an important agricultural product in fresh-cut markets, are sometimes a source of foodborne illness, mainly Salmonella spp. Growth cracks on tomatoes can be a pathway for bacteria, so its detection prior to consumption is important for public health. In this study, multispectral Visible/Near-Infrared (NIR) reflectance imaging techniques were used to determine optimal wavebands for the classification of defect tomatoes. Methods: Hyperspectral reflectance images were collected from samples of naturally cracked tomatoes. To classify the resulting images, the selected wavelength bands were subjected to two-band permutations, and a supervised classification method was used. Results: The results showed that two optimal wavelengths, 713.8 nm and 718.6 nm, could be used to identify cracked spots on tomato surfaces with a correct classification rate of 91.1%. The result indicates that multispectral reflectance imaging with optimized wavebands from hyperspectral images is an effective technique for the classification of defective tomatoes. Conclusions: Although it can be susceptible to specular interference, the multispectral reflectance imaging is an appropriate method for commercial applications because it is faster and much less expensive than Near-Infrared or fluorescence imaging techniques.