Remote sensing image classification with GIS data based on spatial data mining techniques

Data mining techniques are used to discover knowledge from GIS database in order to improve remote sensing image classification.Two learning granularities are proposed for inductive learning from spatial data,one is spatial object granularity,the other is pixel granularity.We also present an approach to combine inductive learning with conventional image classification methods,which selects class probability of Bayes classification as learning attributes.A land use classification experiment is performed in the Beijing area using SPOT multi-spectral image and GIS data.Rules about spatial distribution patterns and shape features are discovered by C5.0 inductive learning algorithm and then the image is reclassified by deductive reasoning.Comparing with the result produced only by Bayes classification,the overall accuracy increased by 11% and the accuracy of some classes,such as garden and forest,increased by about 30%.The results indicate that inductive learning can resolve spectral confusion to a great extent.Combining Bayes method with inductive learning not only improves classification accuracy greatly,but also extends the classification by subdividing some classes with the discovered knowledge.     

Remote sensing image classification with GIS data based on spatial data mining techniques

Data mining techniques are used to discover knowledge from GIS database in order to improve remote sensing image classification. Two learning granularities are proposed for inductive learning from spatial data, one is spatial object granularity, the other is pixel granularity. We also present an approach to combine inductive learning with conventional image classification methods, which selects class probability of Bayes classification as learning attributes. A land use classification experiment is performed in the Beijing area using SPOT multi-spectral image and GIS data. Rules about spatial distribution patterns and shape features are discovered by C5.0 inductive learning algorithm and then the image is reclassified by deductive reasoning. Comparing with the results produced only by Bayes classification, the overall accuracy increased by 11% and the accuracy of some classes, such as garden and forest, increased by about 30%. The results indicate that inductive learning can resolve spectral confusion to a great extent. Combining Bayes method with inductive learning not only improves classification accuracy greatly, but also extends the classification by subdividing some classes with the discovered knowledge.     

Remote sensing activities in Japan

Remote sensing activities in Japan are reviewed under three sectors; the government, private and universities. International activities are also reviewed.

The largest governmental activity is the launch of earth observation satellite by National Space Development Agency (NASDA). Apart from that, remote sensing technology is utilized in many government research institutes.

The first level activities in private companies are research, design and development of rockets, spacecrafts, sensors and receiving stations which are carried out by five big electronic companies. Aerial survey companies and oil exploration companies are presently utilizing technology of remote sensing.

University professors are bearing the responsibility for research and education as well as advising the government and private companies.     

Landscape ecological planning in a Basaltic terrain, central India, using Remote sensing and GIS techniques

In the present study, detailed field survey in conjunction with remotely sensed (IRS-1D, LISS-III) data is of immense help in terrain analysis and landscape ecological planning at watershed level. Geomorphologically summit crust, table top summits, isolated mounds. plateau spurs, narrow slopes, plateau side drainage floors, narrow valleys and main valley floor were delineated. The soil depth ranges from extremely shallow in isolated mounds to very deep soils in the lower sectors. Very good, good, moderate, poor and very poor groundwater prospect zones were delineated. By the integrated analysis of slope, geomorphology. soil depth, land use/land cover and groundwater prospect layers in GIS. 29 landscape ecological units were identified. Each landscape ecological unit refers to a natural geographic entity having distinctive properties of slope, geomorphology. soil depth, land use/ land cover and groundwater prospects. The landscape ecological stress zone mapping of the study area has been carried out based on the analysis and reclassification of tandscape ecological units. The units having minimum ecological impact in terms of slope, geomorphology, soil depth and land use/land cover were delineated under very low stress landscape ecological zones. The units having maximum ecological stress in the form of very high slopes, isolated mounds, table top summits and summit crust, extremely shallow soils, waste lands and very poor groundwater prospects were delineated into very high stress landscape ecological zones. The integrated analysis of remotely sensed data and collateral data in GIS environment is of immense help in evaluation of landscape ecological units and landscape ecological stress zones. The delineated landscape ecological stress zones in the watershed have been recommended for landscape ecological planning for better utilization of natural resources without harming the natural geo-ecosystem of the area.     

遥感与地球系统科学

地球作为一个高度复杂的非线性系统,各圈层(大气、海洋、陆地、生物、冰雪圈、固体地球)尤其是人类活动等任何组成成份的变化,都会引起地球系统的变化。人类可持续发展面临的巨大科学挑战之一是认识人类赖以生存的、复杂变化的地球系统,认识地球系统如何变化及主要驱动因素,认识地球系统未来变化趋势及如何提高对全球变化的适应能力。卫星独特的全球覆盖和日尺度的观测改变了地球科学的研究方法,它强调所能探测到的多时空尺度上的物理动力过程,在全球范围应对气候变化、能源和环境挑战具有重要作用,揭开了地球系统多学科交叉的新纪元。以地球系统的视野,抓住驱动地球系统的关键循环过程(如能量、水、生物化学循环),是当前地球系统科学的发展趋势。地球系统科学(全球变化)研究需要长期稳定、准确性较高的卫星观测数据,以水循环为例,卫星遥感具备获取全球范围水循环关键参数能力,但是系统性综合观测能力不足,整体精确性受到综合化的可靠空间数据集的限制。目前中国正在积极研制发展新型水循环卫星WCOM(Water Cycle Observation Misssion),并寄希望以此为核心传感器发起全球分布式水循环观测星座系统,进一步提高中国在国际水循环观测与地球系统科学研究方面的话语权与领先能力。     

应用Hessian-Affine的遥感图像配准实验

针对存在云雾遮挡、仿射变形的遥感影像,本文提出应用Hessian-Affine与最大信息熵,检测并筛选仿射不变特征,同时选刺同名点估计初始变换参数,对每一个待匹配点预测出一定圆域约束内的对应匹配点集,利用NCC相关系数迭代确定圆域内真同名点,得到初始匹配点集,然后利用均方根误差(RMSE)迭代剔除误匹配,直至完成最佳仿射变换参数估计。结果表明:该算法对发生仿射畸变、气候复杂区域的影像配准表现出较好稳健性和定位精度。