| Due to the investigation, safeguard and rescue use, such as the earthquake relief work, the technology of the Through-Wall-Radar(TWR) imaging has become a hot research these years. The scene imaging technology of building penetrating detection which is based on TWR imaging includes two parts: building layout imaging and hidden targets imaging.TWR imaging can be divided into striped SAR and spotlight SAR according to the detection method. In general, the striped SAR of TWR detects the scene by multi-view and gets a complete scene image by image fusion in the later processing. But due to the complexion of the actual application, it’s hard to fulfil the multi-view detection. And, the intensity difference of multi-layer wall images caused by wall propagation attenuation makes it impossible to express the building layout accurately. What’s more, for synthetic aperture imaging of through-wall-radar, the multi-path propagation between targets and walls introduces the ghosts into the formed image.1. Propose a image enhancement method based on Fuzzy Logic Analysis and Principle Component Analysis, thus suppressing clutter and strengthening the target. The measured data processing verifies the effectiveness of the method.2. Propose a method to compensate the intensity difference of multi-layer wall images caused by wall propagation attenuation, which is based on Hough Transformation. The experimental results validate the proposed method.3. For multi-view images, several image fusion method is presented, such as Threshold Average Weighted Fusion, Adaptive Weighted Fusion, Pixel Value Preferring Fusion and Wavelet Transform Fusion,to form a panorama building-layout image. The above image fusion method is compared by simulation and experiment results.4. Proposes a new algorithm for suppressing multi-path ghosts based on sub-aperture imaging in the condition of single-view detection. The experimental results validate the proposed method.5. In the condition of multi-view detection, different images are multiplied to fuse a final hidden targets image due to the different location of multipath ghosts in different images and the same location of the targets in them. |
