The Key Technology Research On Classification Of Saline Vegetation Based On Object-oriented Method

Soil salinization as a worldwide problem is related to closely related aspects of human life such as environment, resources and ecological problems, and the sustainable development of the human environment also has a lot to do. Although many experts and scholars and even the government propose many methods about the study and governance of salinization, but the land salinization problem not only has not been effectively curbed, and still growing, endangering human and destroy their living environment.Although many plants can not effectively grow because of salinization problems, but the soil of salinization is the favorable environment in which a lot of halophytes can be favorable to life. Therefore, saline vegetation is the indirect indicator reflecting the status of soil salinization, and the vegetation itself is also valuable natural resource. So it is necessary to know its types and distribution, used to guide the development and utilization of saline land resources.On the basis of image segmentation about remote sensing satellite images of the study area, this article considers spectral characteristics, geometric feature, texture feature, other object characteristics vegetation index and salt index, then explores feature extraction and classification system establishment method of remote sensing image based on object-oriented classification method, and identifies the categories of saline vegetation in the study area, the last obtains classification and distribution information of various saline vegetation resources in the study area, and realizes high precision drawing of natural distribution regions about saline vegetation; through contrasting, analyzing and evaluating with the results of the traditional classification method in the study area, further elaborates object-oriented image classification method is superiority to traditional image classification method based on the pixels about the saline vegetation classification processing.This paper intends to solve the key problem applying the object-oriented classification method to the classification and recognition of the saline vegetation, explore image segmentation method and optimal segmentation scale which are suitable for saline vegetation classification and select the optimal characteristics combination of classification and recognition comparing with the traditional classification method based on pixels.This study provides a scientific basis for scientific research and effective governance of salinization regional resources, can be used to guide the development and utilization of saline land resources, is advantageous to the ecological sustainable development and protection of natural resources in the Yellow River Delta.