The Study of Consistency Imaging Method for Optics and SAR Based on Common Aperture

Optical and radar images suffer from poor image interpretation in most existing non-common aperture regimes. To solve this problem, it is proposed to unify the optical imaging system and radar imaging system in a common aperture regime, ground targets in the same area can be imaged at the same time and the same angle, it can avoid the inconsistencies in observation time, angle, and range that exist in the non-common aperture regime. Considering the optical imaging principle of the optical frame camera, and its central projection properties, Synthetic aperture radar (SAR) is side-view imaging and SAR imaging relies on platform motion, thus constructing a consistency imaging model for optical and SAR under the common aperture. At the signal level, real-time alignment of optical and SAR images is achieved, the results of SAR imaging are aligned with the optical image, which simplifies the complex process where imaging and alignment are normally separate. Based on the geometrical model of optical centre projection imaging and the imaging principle of SAR based on distance information, the consistency imaging model of optics and SAR under common aperture is constructed, which achieves the signal level alignment. In the common aperture regime, the camera is shooting sideways to get a tilted photo, solves the position information of the target from the optical photo, and corrects the tilt of the solved target area. By linking the central projection geometry of optics with the pulse-by-pulse and pixel-by-pixel signal processing of the time-domain BP algorithm in SAR, after solving the position of the target from the optical photograph, a high-resolution SAR image can be obtained by using BP algorithm. The alignment accuracy of SAR imaging and optical images is better than one pixel, without the need for complex alignment processing in the image domain. With the integration of optical camera and SAR imaging in the common aperture regime, the proposed consistency imaging model can generate two automatically aligned images, which simplifies the data processing flow of conventional image alignment, the separation of traditional imaging and registration is effectively broken through, and improves the processing efficiency, also facilitates the image fusion and target detection at the same time.