Research On Multiband Fusion Imaging Technology

Multiband fusion imaging is an emerging radar imaging technology. It focuses on the fusion of observation information from mutiband radars so as to improve the resolution. It is also in favor of target recognition and image comprehension. The information fusion based on the multi-radar observations can obviously improve the radar imaging resolution and the scattering center parameters estimation precision. Aiming at the background of multiband spectra fusion of fast moving targets, this dissertation studies some pivotal problems including wideband radar echo model of moving targets, coherent processing, multiband fusion imaging.Firstly, the traditional multiband fusion imaging processing was improved on the fusion imaging quality.It is the theoretical basement of coherence and fusion algorithm to build the wideband radar echo model of moving targets. The model when the transmitting signal was linear frequency modulate (LFM) signal and the frequency response was geometry theory of diffraction (GTD) model is given in the second chapter; based on which two predisposal process using matched filtering (MF) and dechirp processing. After predispose we can get the subband frequency spectra.Coherent processing is an important step before multiband fusion. The third chapter analyzes the parameters which cause the uncoherence among multibands and based on which a coherent function under the condition that the target moves fast is presented. A method of frequency-domain coherent method based on genetic algorithm (GA) was presented to solve the weak point of traditional optimization algorithm which is inclined to trap in local minimum. And the GA utilizing the middle overlapped spectra is given to reduce the error caused by spectra extrapolation. But the problem that the GA causes heavy computer load also exists and GA need to extrapolate the spectra of subbands. So the author presents a method using the poles'information directly, further more the coherent method in time domain that doesn't need to extrapolate the spectra is put forward. Finally, we compare the performance through computer simulation.Fusion imaging is the most important step in the whole process. The fourth chapter introduces the basic principle. Based on data rearrangement and spatial smoothing techniques the covariance matrix that has the property of translation invariant was built. By minimum description length (MDL) the rank of model is estimated. Then the method to estimate the signal poles is presented. Thus the location, frequency dependent coefficient and amplitude of scattering centers are estimated by poles'information. Finally, Montcarlo simulation as well as simulation combined with transmitting signal was carried out to prove the above method. Microwave darkroom experiment is an effective method to test the efficiency of multiband coherent process and fusion imaging algorithm. The fifth chapter discusses the simulation method of uncoherence factors, the experiment environment and the steps of experiment. The real data is dealt with and the result proves the efficiency of the presented method.