In order to overcome the 1imitation of traditional signal processing methods used in seismicreflection signal which is a kind of nonstationary signa1, we design wavelet noise-eliminationmethods by using the different propagation properties of singular signal and random noise inmultiscale space.

First,the multi-scale space of the image is created via a multi-level box filter,and then the feature points are detected in scale space by considering.

According to the multi-scale space representation of an image,the reference image and real-time image were decomposed to different scale space sub-images as well as providing rich image features by nonlinear critical filter.

Habitat Selection and Prediction of the Spatial Distribution in Greater Horseshoe Bats, Rhinolophus Ferrumquinum at Multiple Spatial Scales

Study on automatic interpretation of remotely sensed imagery based on multi-information space;

Research on Integration of Multi-source Spatial Data;

This paper discusses the existing problems of spatial data sharing and integration,puts forward a model of multi-source spatial data integration based on grid,analyses the key techniques and architecture of this model,and introduces the prospect of spatial data integration based on grid technology.

This article elaborates the production and performance of multi-source spatial data as well as the hindrance of multiple data forms to GIS development,and discusses multi-source data intergration methods.