Imaging as a new function of radar is widely used, which has the characteristic of all-weather, day/night and long range, enhance radar's information acquisition capability, especially battlefield awareness ability. Now, most of radar in abroad that needs high battlefield awareness ability has the 2-D imaging function, which has great value in both civilian and military applications.Radar imaging can be divided Inverse synthetic aperture radar (ISAR) imaging and synthetic aperture radar (SAR) imaging. Radar imaging is a practical technique, a huge number of raw data must have to study it. In the past, because the number of wideband radar is few in our country, so the raw data is very rare, most of researches are based on simulation and raw from abroad, that restricts further researches. In 1993, a wideband ISAR radar with 400MHz bandwith is developed on the support of "863" program, and many raw data (or real data) is collected, ISAR research becomes more deep by providing these data to many colleges, universities and institutes for research. In these years, many research institutes, such 607, 38,14 and 206, have developed airborne SAR system, and have collected many raw data from many times test-fly. This dissertation study the radar imaging methods based on real data collected by our country on the basis of existing research work. The following is the summarization of the main work: Chapter 2 studies the principle of ISAR imaging for maneuvering targets. ISAR and SAR all utilize the relative motion of target (scene) and radar to improve cross-range resolution, and then obtain the target's image. Usually, in SAR, the radar moves steadily by controlling, and the deviation can be measured, but, in ISAR, the target's motion is non-cooperative, and is difficult to measure. When target maneuvers, and if we take target as reference, radar will form a complicated inverse synthetic aperture (array) manifold equivalently in space, this chapter studies the imaging problem systematically in this situation. Chapter 3 studies the translational motion compensation (TMC) and imaging methods of maneuvering target. We first discuss TMC, which is usually discomposed into two step: envelope alignment and autofocus. The existing envelope alignment algorithms are still effective for not too big maneuvering targetPhD Thesis of Xidian Universitywhose migration through resolution cells (MTRC) is not take place, but the existing autofocus algorithms are not effective for maneuvering target in practice and in theory according to coherent summation principle. So we propose an iterative coherent summation autofocus (ICSA) algorithm, PGA is a particular case of ICSA. Then we discuss maneuvering target imaging problem, in fact, which is a instantaneous spectrum estimation problem, most existing algorithms are only effective when scatterers' echoes are linear frequency modulation (LFM) signals, for the situation when scatterers' time-frequency distribution is not linear, we put forward a adaptive chirplet decomposition method to estimate instantaneous frequency and instantaneous complex amplitude of multi-component polynomial phase signals, and propose a fast adaptive chirplet decomposition imaging (ACDI) algorithm by utilizing "Clean" technique. Real data processing proves the proposed ICSA algorithm and ACDI algorithm are effective. Chapter 4 studies the IS AR imaging of ship. Many researches of IS AR imaging for airplane have been done in our country, ISAR imaging for ship is also very important in national defense, but the research for it is very few. Although the difference is little in principle between airplane imaging and ship imaging, the pose variation of ship is more complicate for the fluctuation of sea, so it need to be studied specially. This paper propose a Range-Instantaneous Doppler Algorithm (RIDA) based on multi-component amplitude modulation and linear frequency modulation (AM-LFM) signal parameter estimation method to obtain good ship ISAR images from real data. Chapter 5 investigates the properti