Analysing decadal land use/cover dynamics of the Lake Basaka catchment (Main Ethiopian Rift) using LANDSAT imagery and GIS

Development of accurate classification methods for rapidly changing catchments like that of Lake Basaka is fundamental to better understanding the catchment dynamics, which were not addressed in previous studies. Thus, the aim of this study was to map the decadal land use/cover (LUC) regimes of the Lake Basaka catchment, utilizing time series of LANDSAT images and to analyse the changes that occurred at different time periods. Both unsupervised and supervised image classification systems were utilized in Earth Resources Data Analysis System (ERDAS) Imagine (9.1). Appropriate pre‐ and postprocessing also was utilized. Seven major LUC classes were identified in the final land cover maps produced after the supervised (maximum likelihood) classification exercise. The analysis results indicated the Lake Basaka catchment had experienced a drastic change in its LUC conditions over the last 4–5 decades because of rapid increases in human settlement, deforestation, establishment of irrigation schemes and Awash National Park (ANP). Approximately 18 924 ha of forest and 4730 ha of grazing lands were devastated between 1973 and 2008. At the same time, there was a shift in land cover from forests/woodlands to open woodlands, shrub and grazing lands. The land cover classifications generally were achieved at a very high overall accuracy (84.34%) and overall kappa statistics (0.802), substantiating the value of using the classified LUC in this study as an input to hydrological models. This study results provide an opportunity to better understand and quantify the hydrological response regimes of the lake catchment from the perspective of changing LUC conditions during different hydrological periods and the resulting dynamics of the lake water balance.     

太湖蓝藻信息采集系统的建设与应用

太湖蓝藻信息采集系统项目的建设,是结合卫星遥感、3G图像传输、浮台水质监测及GIS等技术监测太湖蓝藻。同时项目中采用的3G视频传榆技术,为今后水利水文信息化项目中条件受限制的站点上传图像信息也起示范作用。     

结合RSTC不变矩和LVQ神经网络的红外图像目标识别方法

通过分析比较红外图像的目标特征,为了达到理想的识别效果,在Maitra不变矩的基础上进行优化,选取RSTC不变矩作为目标识别的特征向量.采用LVQ神经网络建立识别模型,充分发挥神经网络的智能优势.对采集到的红外图像进行了测试实验,结果表明该方法可以提高识别效率.     

基于CBERS-02B高分辨率遥感影像分割的最优尺度选取

高分辨遥感影像的信息提取对矿山的监测有着重要意义,影像分割是高分辨率遥感影像信息提取关键的一步．文中以赣州稀土矿山的高分辨率遥感影像为源数据,采用分形网络演化多尺度分割算法对高分辨率遥感影像进行分割,通过Robert算子的边缘提取结果计算分割参数获取分割尺度,并在此基础上设置不同分割尺度对影像进行分割,对比分析分割结果表明,沉淀池的最优分割尺度为35．最后利用分割质量值对分割结果进行评价分析,验证了试验结果的正确性．     

基于主色偏差矩阵的图像检索方法

颜色特征是图像检索领域中最为重要的视觉特征之一.对于基于颜色的图像检索系统,其性能在很大程度上取决于颜色特征的描述和匹配方法的有效性.本文使用主色偏差矩阵来描述颜色特征,并使用相关系数进行相似性度量.经实验证明,该方法对于图像的几何变换和空间变换,具有较高的检索查全率.     

基于柱镜光栅LCD的裸眼显示技术研究

为了获得更高质量的目标图像,克服现有方法的不足,提出基于柱镜光栅LCD显示设备的光栅参数测量、立体图像合成与分解的改进方法.     

Novel method for haze degraded image scene rendition

To remove haze rapidly and effectively, equivalence between Retinex model and Monochrome Atmospheric Scattering model is explored, and then a novel solution for the Monochrome Atmospheric Scattering model is proposed based on the variational Retinex. Firstly, a variational model for the atmosphere transmission map is constructed, so the transmission map estimation problem can be formulated as a Quadratic Programming optimization problem. Secondly, we calculate the transmission map and resolve the Monochrome Atmospheric Scattering model in each channel to get the Reflectance image; Finally, the Color Constancy is adopted to restore the color of the dehazed image. Experimental results show that the proposed method could remove haze rapidly and effectively, and has some advantages over other algorithms.     

Efficient algorithm for mismatching point elimination with interzone limited pre-testing

The elimination of mismatching point is a key step in image mosaic. To speed up the mismatching point elimination in image matching, an efficient algorithm for mismatching point elimination is presented. Based on the RANSAC(random sample consensus) algorithm, the algorithm aims at the feature of the image sequence. First, the match points are sorted, divided into three parts and piecewise picked randomly to estimate the transforming matrix. Second, the matrix is cursorily checked with a novel interzone limited pre-test model, a further check is executed on the matrix that passes the pre-testing to get inliers. Finally, the least-square method is used to the inliers to get the real matrix. Experimental results suggest that the proposed algorithm is of lower complexity, higher accuracy and stableness especially in cruel conditions, which meets the demand of image sequence mosaic well.     

基于姿态定位算法的水下电缆监测系统

针对海洋科学考察中连接设备的水下电缆无法观测的现状,该文提出了一种基于姿态定位算法的水下电缆监测系统。系统采用两片V,11SCA60单轴倾角传感器对电缆不同位置处的双轴倾角数据进行采集,并通过CAN总线传送到水下主控CPU,数据经打包后由同轴电缆或光电复合缆传送至甲板上位机,通过姿态定位算法在上位机中完成电缆姿态监测。实验证明该系统可准确地定位电缆姿态并能够对危险操作和异常状态及时做出预警,为设备水下投放提供保障。