| Synthetic aperture radar(SAR)has been widely applied to military and civil fields for it can provide 2-D high resolution image of the observed scene in day and night under all weather conditions.For the spaceborne SAR,a resolution of several meters can meet application requirements in the military field of battlefield environmental monitoring,military deployment and in the civil fields of glacier detection,vegetation measurement,soil detection,earthquake monitoring and so on.At this time,the swath width of spaceborne SAR is particularly important,which is related to the SAR's wide area coverage,and whether SAR can complete the observation and reconnaissance of the wide area scene in a short time.Spaceborne SAR imaging with wide swath and target detection technology has become an important research direction of SAR imaging.For the development requirements,this dissertation studies some key points on the imaging and target detection algorithms of spaceborne SAR.The whole dissertation mainly contains four aspects as follows:1.Aiming at the situation that the Doppler ambiguities varies with the baseband frequency and the steering vectors are unknown,an unambiguous reconstruction method based on Doppler spectral estimation is proposed.The algorithm obtains the aliased Doppler spectral structure by means of Capon spectral estimation,and then number the ambiguity components and constructs the steering vector to realize the unambiguous reconstruction.Then,the reconstruction performance of the algorithm is analyzed.When using the constant number of ambiguity components for reconstruction,it always introduces too much noise and reduce the signal to noise ratio(over-reconstruction)or causes the loss of signal spectrum and reduces the azimuth resolution(under-reconstruction).However,this method can maintain the integrity of the Doppler spectrum and increase the signal to noise ratio of the reconstructed signal.The algorithm is an improvement and extension of existing algorithms,which can be used as a pre-processing step combined with existing methods.The algorithm can estimate the azimuth antenna pattern,which can provide basis for further calculation of Doppler center,azimuth ambiguity to signal ratio and other system parameters.Problems that may occur in practical applications are put forward in this chapter.2.According to the principle of azimuth phase coding(APC),a range phase coding waveforms is proposed,which can distinguish the range ambiguous signal and the target signal in range frequency domain.Then the range ambiguous energy can be suppressed by using filtering.Since the suppression effect of APC method is limited caused by the low oversampling rate,a novel range ambiguity suppression technique based on a two-dimensional phase coding is proposed.By transmitting two-dimensional phase coded signals and demodulating the received signals,the two-dimensional spectrum of the range ambiguity is shifted along both range and azimuth directions.Considering the range frequency and Doppler oversampling,the spectrum of the range ambiguity located outside the data region of spectral support can be filtered out,which suppresses the range ambiguity.This method removes the range ambiguous signal energy not only in Doppler frequency domain but also in range frequency domain.It allows for a simple signal processing and provides more degrees of freedom in choosing the pulse recurrence frequency.Moreover,the method is effective against not only the ambiguous energy of point targets but also that of distributed targets.Finally,the chapter shows the performance of ambiguity suppression.It also makes a detailed analysis of the results in the aspect of the signal energy.which means that the method can decrease of the range ambiguity to signal ratio(RASR)and upgrade of the SAR image quality.3.Since the traditional cell averaging constant false alarm rate(CA-CFAR)algorithm has the low computational efficiency in calculating the moment estimation of the background clutter,cell averaging based on convolution constant false alarm rate(CCA-CFAR)for target detection in SAR images is proposed in this paper.As a first step,the statistic matrices of the SAR image are computed according to the background clutter distribution model.Then,a convolution model is built for the CA-CFAR detector to realize the moment estimation,and all pixels' background clutter distribution function can be obtained.Finally,calculate the detect threshold for each pixel and detect whether the pixel is the target point.The computational complexity of the proposed CCA-CFAR algorithm is analyzed in detail,and the amount of calculation for each step is given.Then the performance of this algorithm is analyzed by simulation experiments,which is determined by the size of the SAR image,the background clutter data window and the protection window.The algorithm has the advantages of low complexity and high computational efficiency,and can achieve SAR image real-time target detection.It is an improvement and extension of the existing algorithms,which can also be used as the pre-detection method combined with existing methods.4.Aiming at the application requirement of ultra wide swath SAR imaging,a new circular scanning SAR working mode is proposed.By using circular scanning of the antenna,the imaging scene forms a wide annular imaging area,which realizes ultra wide swath SAR imaging.In this chapter,a geometric model of the circular scanning SAR is built according to the beam scanning.Then,the optimal swath mode and the optimal resolution mode are proposed.The system parameters of the circular scanning SAR under the above modes are designed and demonstrated.Finally,optimal system parameters in the optimal swath mode and optimal resolution mode are obtained through analysis,which provides the basis for the engineering application of the circular scanning SAR. |
PDF Full Text Request