| Because HRRP(high-Resolution Range Profile)can provide more target structure and shape information, research on Space Target Recognition based on Range Profile is of significance for both theory and practice. Discussions on technologies of Range Profile Preprocessing, Feature Extraction and Classification have been made in this thesis. The main contents and ideas include:1. Research carried out on Space Target Range Profiles Preprocessing and Feature Extraction has been discussed in the thesis.HRRP has the following sensitizations: azimuth, shift and amplitude, to avoid these effect, range grid, amplitude normalization and average profile has been adopted. Based on Preprocessing, scatters position and some amplitude feature is extracted in this thesis, including Central Moments Feature, Range-Domain Feature, Amplitude-Domain Feature and Wave-Fluctuation Feature.2. Research carried out on Range Profile ClassifierIn this thesis, the description of research on Support Vector Machine based on Statistical Learning Theory has been made. Research on three kinds of satellite data,'one-versus-one'based on Multi-classification SVM has been adopted. Because of correlation among these features, different combined features has different experimental results; Compared with KNN Classifier, SVM has a more optimal performance.3. Research carried out on Space Target recognition based on Multi-sensors fusionThe characteristic of analysis on Multi-sensors has been made in this thesis, including electromagnetic scattering characteristic, infared characteristic, Orbit characteristic and optical characteristic. This paper puts forward Multi-Stage Recognition combined with Multi-sensors characteristic. This method can discriminate target attribute with the information of Multi-Sensors sufficiently.Summaries are made and prospects of the Space Target Identification of Range Profile and Multi-Sensors fusion Identification have also been put forward in the final part of the thesis. |