Surface chlorophyll a estimation in the Arabian Sea using IRS-P4 Ocean Colour Monitor (OCM) satellite data

The IRS-P4 satellite launched on 26 May 1999 by Indian Space Research Organisation (ISRO) carried an Ocean Colour Monitor (OCM) payload, primarily designed to measure ocean colour, the spectral variations of waterleaving radiance that can be related to the concentration of phytoplankton pigments, coloured dissolved organic matter and suspended particulate matter. Remote sensing data obtained from IRS-P4 OCM sensor is processed for the removal of atmospheric effects of Rayleigh and aerosol scattering to derive normalized water-leaving radiance in visible channels of 412, 443, 490, 512 and 555 nm. Ocean chlorophyll-2 (OC2) global bio-optical algorithm was used to convert normalized water-leaving radiance data to chlorophyll a concentration. For quantitative accuracy assessment, OCM-derived chlorophyll a concentration values were compared to the in situ measurements of chlorophyll a, conducted during ORV Sagar Kanya SK-149c ship cruise during 15-27 November 1999 in Arabian Sea. In general OCM-derived chlorophyll a showed a good correlation ( r 2 =0.90, N = 8, RMS=0.125 mgm -3 ) with measured chlorophyll values.     

Phytoplankton distribution in the Arabian Sea using IRS-P4 OCM satellite data

For rapid static mode global positioning system (GPS) satellite surveying, our data processing method uses linear combinations of dual-frequency carrier phase observables, which are double differencing wide lane signals between the satellites and the receivers. Wide lane ambiguily parameters estimated in a least squares sense lead to corresponding candidate integer ambiguities on LI carrier waves. This shortens the search time appreciably in order to resolve the integer ambiguities. A contrast statistic and an ambiguily function provide indices which are utilized together to guarantee the correctness of the relative position solutions. Coordinate repeatabilities obtained are qualified for the generation of local ground control points. With five vertical control points, geometric heights from GPS surveying are transformed into orthometric heights. They are then compared to heights independently derived from precise levelling. Accuracies in height ranging from ± 0·8 to 1·6 cm are realizable for an experiment area of about 4 km by 5 km.     

Change analysis using IRS-P4 OCM data after the Orissa super cyclone

The Orissa super cyclone hit the eastern coast of India in October 1999 and resulted in drastic changes to land and coastal water areas. This Letter demonstrates the use of IRS-P4 Ocean Colour Monitor (OCM) data for mapping changes resulting from this cyclone. The results show pronounced change in the vegetation cover of Orissa province and also change in the chlorophyll and suspended matter concentrations of the ocean water. The eight spectral channel capability of OCM sensors enabled us to map land cover changes and changes in coastal water after the cyclone.     

Simulations for optimal payload tilt to avoid sunglint in IRS-P4 Ocean Colour Monitor (OCM ) data around the Indian subcontinent

Ocean Colour Monitor (OCM) payload, onboard Indian Remote Sensing Satellite (IRS)-P4, detects the water constituents from the spectrum of solar radiation backscattered from the ocean waters. The radiation received by the sensor is contaminated by the specularly reflected solar radiation from the water surface. This specularly reflected radiation, called sunglint, contains no information on the water constituents, as it has not entered into the seawater and interacted with it. The intensity and spread of sunglint is determined by the solar illumination and sensor viewing directions and the sea surface roughness caused by the wind. For the accurate estimation of oceanic constituents, it is essential to minimize the sunglint in the detected radiances (preferably below ～2-3%). In this work, sunglint simulations were computed for the instrument specifications of OCM and the optimal sensor viewing tilt angle identified for each month for the oceans around India.     

基于整体最小二乘法的高分辨率图像重建算法

提出一种用于高分辨率图像重建的整体最小二乘算法。在现有多数重建算法中,假设系统矩阵是精确的而误差主要源于采样图像,但实际上抖动误差也出现在系统矩阵中。该方法能同时最小化这两种误差,采用基于正则化的Rayleigh商来光滑解,用共轭梯度算法来迭代求解该正则化Rayleigh商的最小化函数。实验证明该方法对于抖动系统矩阵是稳定和精确的。     

Cloud screening in IRS-P4 OCM satellite data: potential of spatial coherence method in the absence of thermal channel information

Cloud screening of satellite data for the remote sensing of atmospheric aerosols, ocean sediments, chlorophyll, and phytoplankton in the marine environment is a major problem in the absence of information from thermal channel. This is particularly the case with the data from some of the highly potential satellite sensors such as the Ocean Colour Monitor (OCM—on board the Indian Remote Sensing Satellite, IRS-P4) and the SeaWiFS. Two main tests conventionally used for cloud screening of data from such satellite sensors are the threshold method applied to visible and near-IR bands and the visible to near-IR channel ratio method. These methods do not have the potential to eliminate the pixels with small cloud fractions, leading to overestimation of the aerosol optical depth (AOD) derived from satellite data, and might also identify the pixels with high values of AOD as cloudy. The purpose of this paper is to study the potential of Spatial Coherence Test (SCT) applied to the data from the near-IR bands for cloud screening of satellite data over the oceanic environment. We use here the data from IRS-P4 OCM. Though more computationally intensive, the SCT does not suffer from the serious limitations of the threshold and channel ratio methods and is found to be superior in identifying the clear sky pixels that are not affected by clouds. Although the SCT applied to near-IR channel data may be overestimating the number of cloud affected pixels, it neither leads to overestimation of AOD nor identifies the pixels with high AOD values as cloudy.     

基于高分辨率SAR图像的道路提取算法

针对合成孔径雷达(synthetic aperture radar,SAR)图像中道路边缘的特点,提出了一种基于多条件加权法的高分辨率SAR图像道路提取算法。该算法使用FROST滤波抑制相干斑噪声,并使用OSTU算法对高分辨率SAR图像进行二值化,对二值化后的SAR图像进行膨胀与腐蚀,再使用五邻居边缘检测器与多条件加权法提取道路的一个边缘,最终使用桥连接模式提取出完整的道路边缘。实验结果表明,该算法可以消除噪声,消除障碍物的干扰,有效的提取道路边缘。     

高分辨率遥感图像小波融合算法研究

针对现有融合方法存在的光谱信息和空间细节信息不能较好兼顾的问题,提出了一种遥感图像小波融合算法。根据高分辨率遥感图像自身的空间尺度,采用多进小波分析技术较好地解决了不同尺度遥感图像空间细节的表达问题,同时利用基于方向能量统计量的自适应融合准则较好地解决了融合中的光谱信息保留问题。实验结果表明,该融合方法在保持全色图像的高空间分辨率的同时,提高了融合图像的光谱保持度。     

基于IHS的高分辨率遥感影像自适应融合算法

在分析几种IHS算法的基础上,提出了一种适合于高分辨率遥感影像的自适应加权融合算法。该算法基于快速IHS变换,从亮度成份构成和空间细节注入程度两个方面加以改进,采用相关系数来调整权重,避免了人工指定权重带来的主观因素偏差,在提高影像空间细节和光谱保持之间取得了较好的均衡,同时还具备了高计算效率。通过融合实验并与几种改进的IHS算法的对比分析,该算法均优于其他几种改进算法。     

高分辨率可见光图像引导红外图像超分辨率的Transformer网络

目的 红外图像在工业中发挥着重要的作用。但是由于技术原因,红外图像的分辨率一般较低,限制了其普遍适用性。许多低分辨率红外传感器都和高分辨率可见光传感器搭配使用,一种可行的思路是利用可见光传感器捕获的高分辨率图像,辅助红外图像进行超分辨率重建。方法 本文提出了一种使用高分辨率可见光图像引导红外图像进行超分辨率的神经网络模型,包含两个模块：引导Transformer模块和超分辨率重建模块。考虑到红外和可见光图像对一般存在一定的视差,两者之间是不完全对齐的,本文使用基于引导Transformer的信息引导与融合方法,从高分辨率可见光图像中搜索相关纹理信息,并将这些相关纹理信息与低分辨率红外图像的信息融合得到合成特征。然后这个合成特征经过后面的超分辨率重建子网络,得到最终的超分辨率红外图像。在超分辨率重建模块,本文使用通道拆分策略来消除深度模型中的冗余特征,减少计算量,提高模型性能。结果 本文方法在FLIR-aligned数据集上与其他代表性图像超分辨率方法进行对比。实验结果表明,本文方法可以取得优于对比方法的超分辨率性能。客观结果上,本文方法比其他红外图像引导超分辨率方法在峰值信噪比(pea...     

Comparison of chlorophyll distributions in the northeastern Arabian Sea and southern Bay of Bengal using IRS-P4 Ocean Color Monitor data

Water masses of different temperature and salinity partially govern the distributions of chlorophyll concentration. In the present paper, chlorophyll concentration has been deduced over the northeastern Arabian Sea and the southern Bay of Bengal using Ocean Color Monitor (OCM) data onboard Indian Remote Sensing Series Polar satellite (IRS-P4). The chlorophyll concentration is found to be higher in the northeastern Arabian Sea compared to those in the southern Bay of Bengal. The higher chlorophyll concentration is found during the northeast monsoon compared to those during the pre- and post-monsoon period and also in the coastal water compared to the open ocean. Higher spatial and seasonal chlorophyll distributions in the northeastern Arabian Sea compared to the southern Bay of Bengal is attributed to the shallow depth of the shelf in the Arabian Sea.     

Binning schemes for partition-based compound selection

Partition-based approaches to the selection of structurally diverse sets of compounds involve allocating compounds to the individual elements of a multidimensional grid that spans the available chemical space. The space is defined by an appropriate set of chemical properties, with subranges of the values of these properties being used to define the constituent elements, or bins. This article compares several binning schemes in terms of their ability to provide an even distribution of compounds across the available space and to maximise the numbers of active molecules identified in simulated assay experiments.     

Wavelet-based image registration technique for high-resolution remote sensing images

Image registration is the process of geometrically aligning one image to another image of the same scene taken from different viewpoints at different times or by different sensors. It is an important image processing procedure in remote sensing and has been studied by remote sensing image processing professionals for several decades. Nevertheless, it is still difficult to find an accurate, robust, and automatic image registration method, and most existing image registration methods are designed for a particular application. High-resolution remote sensing images have made it more convenient for professionals to study the Earth; however, they also create new challenges when traditional processing methods are used. In terms of image registration, a number of problems exist in the registration of high-resolution images: (1) the increased relief displacements, introduced by increasing the spatial resolution and lowering the altitude of the sensors, cause obvious geometric distortion in local areas where elevation variation exists; (2) precisely locating control points in high-resolution images is not as simple as in moderate-resolution images; (3) a large number of control points are required for a precise registration, which is a tedious and time-consuming process; and (4) high data volume often affects the processing speed in the image registration. Thus, the demand for an image registration approach that can reduce the above problems is growing. This study proposes a new image registration technique, which is based on the combination of feature-based matching (FBM) and area-based matching (ABM). A wavelet-based feature extraction technique and a normalized cross-correlation matching and relaxation-based image matching techniques are employed in this new method. Two pairs of data sets, one pair of IKONOS panchromatic images from different times and the other pair of images consisting of an IKONOS panchromatic image and a QuickBird multispectral image, are used to evaluate the proposed image registration algorithm. The experimental results show that the proposed algorithm can select sufficient control points semi-automatically to reduce the local distortions caused by local height variation, resulting in improved image registration results.     

一种新的数字图像放大4／3倍快速算法

提出一种新的数字图像放大4／3倍快速算法。该方法既解决了零阶插值算法的棋盘格效应问题,又在运算效率上比传统的一阶插值算法提高了90％。     

四阶高分辨率熵相容算法

针对一维Burgers方程和一维Euler方程组的数值求解问题,提出了一种四阶高分辨率熵相容算法。新算法时间方向采用半离散方式,空间方向应用四阶中心加权基本无振荡(CWENO)重构方法,数值通量引入Ismail通量函数,将新的四阶算法应用于静态激波问题、激波管问题以及强稀疏波问题的数值求解中,并将所得结果同准确解以及已有算法所得结果进行了分析与比较。数值结果表明:新算法计算结果正确、分辨率高,能够准确捕捉激波及稀疏波,并能有效避免膨胀激波的产生。新算法适用于准确解决一维Burgers方程和一维Euler方程组的数值求解问题。     

共振算法用于图像分割

计算机视觉和识别技术在智能控制中的作用越来越重要.通过对利用共振理论进行图像分割方法的研究,提出了一个图像分割的实用算法,这通常是控制一个智能系统的第一步.共振算法是一种无教师学习算法,在门限的限定下用来将相邻的相同或相似的特征向量(或像素)组成同一特征空间(或区域).此算法在一定程度上容忍图像分割中出现的纹理特征强度逐渐变化现象.实验结果验证了共振算法是强调相邻特征相量(或像素)的相似性,而不是试图通过一个针对全局的门限来分割图像.     

一种基于仿射变换的SURF图像配准算法

传统的SURF算法对仿射变化较大的图像配准效果差。为此,提出了一种仿射-加速鲁棒性特征(Affine-SURF)的图像配准算法,通过增加经度角和纬度角不变特征引入仿射形变参数来模拟图像在不同角度的变形。实验结果表明,与SIFT、SURF、MSER等配准算法相比,该算法能够获得更多特征匹配对,提高了算法对仿射变化的鲁棒性。     

一种新的DOA估计的高分辨率算法

波达方向估计（DOA）在阵列信号处理中非常重要,而传统算法如多重信号分类算法（MUSIC）需要对阵列接收数据的协方差矩阵进行特征分解,并在全空域进行谱峰搜索,运算量巨大,尤其是二维DOA估计算法还存在稳健性较差的问题。提出了一种基于最小互熵谱分析的DOA算法,该算法只需要一次较少阵元的阵列采样（快拍）数据即可得到具有高分辨率的谱分析结果。进一步采用倒谱法,通过逆FFT变换提高了最小交叉熵谱分析算法的收敛速度。数值仿真结果表明,该算法较常规空间谱估计方法有更高的分辨力和更小的运算量,能够对阵列信号进行实时处理。算法不依赖于预先估计的信源数目,具有较好的宽容性,较高的分辨力以及极低的旁瓣电平。     

改进的4D-SPIHT医学图像无损压缩

在3D-SPIHT编码的基础上,提出基于不对称小波树的分方向4D-SPIHT编码算法,通过构造4维不对称小波树结构,使得每一维上的小波分解级数可以灵活选择;根据小波树特点,将各小波频带按方向独立进行SPIHT编码,从而加快编码速度。实验结果表明,该算法能较有效地去除三维体数据之间的相关性,在不明显增加算法复杂度的基础上,提高压缩性能和编码效率。