The Research On Wheel Load Dynamic Detecting Of Weigh-in-Motion Systems

In the background of the rapid development of domestic high-resolution images,it is more and more important to improve the utilization rate of information resources in domestic high-resolution images.The remote sensing image information extraction the most efficient and effective way,is the image of high-precision classification.After domestic and foreign scholars continue to study,image classification method has been very rich and perfect.In order to make these mature classification methods can be effectively applied to the domestic high resolution image,it is very important to study the applicability of remote sensing image classification in China.In this paper,the Landsat-8 images,domestic GF-1,ZY-3 and comparative analysis were selected as the object of study in Changchang Expressway,Hunan Province.After basic preprocessing and image fusion,the remote sensing images a study on the applicability of classification.The main contents and conclusions are as follows:(1)The fusion effect of the domestic high resolution remote sensing image fusion method is studied,and the evaluation method of the subjective and objective aspects is selected,and the best effect NNDiffuse method is used to fuse the image of the study area.(2)The object-oriented classification of the best segmentation scale study,compared with the scale of the segmentation effect,made the remote sensing image object-oriented classification method is divided into 50 appropriate scale.(3)The applicability of the remote sensing image classification method was studied.The traditional supervised classification was the highest in the classification of domestic high resolution images represented by GF-1 and ZY-3.Object-oriented classification in the study area ZY-3 Image classification is better than supervised classification.Non-supervised classification is low in both Landsat-8 images and abroad,and does not apply to the classification of remote high-resolution remote sensing images.(4)The influence of image fusion on the classification accuracy is studied.In general,the classification accuracy of the image is improved after the fusion processing,and the high resolution image improves the classification accuracy compared with the low resolution image Sensitive.In this paper,the applicability of each classification method in the domestic high resolution remote sensing image is taken as the research focus.The research contents and results can be used as the basis of high resolution remote sensing image classification method such as high resolution and resource series in China.The preprocessing process(especially image fusion)provides an empirical reference for object-oriented classification and scale setting,and can also promote the application of domestic high-profile remote sensing image resources in land use dynamic monitoring and roadfield vegetation information extraction.