The Application Of Anti-symmetrical Mapping Transform Technology In Researching Scene Matching Correlator

Imaging guidance has been a primary development trend of precise guidance weapon, and been thought to be more and more important. It contains infrared imaging guidance, visible light imaging guidance and SAR image guidance and so on. Its key equipment is scene matching correlator, and its key technology is scene matching. It realizes independent navigation and guidance by matching real-time image and reference image. Simultaneity, as to imaging guidance, large sight, high resolving power and real-time processing are its development trend. Large sight and high resolving power must result in data's increase greatly, which increases the difficulty in realizing real-time processing.For the sake of resolving the conflict, anti-symmetrical mapping technology is applied to scene matching correlator. It anti-symmetrically maps or samples symmetrical image or scene. On the one hand, it can make sure high resolving power sampling the central region, on the other hand, as the distance away from the focus center increases, the resolving power will decrease, and data will be compressed at the peripheral region, which can reduce operation, and improve real-time processing. Simultaneity, based on effective attention transfer mechanism, the focus center is confirmed by selecting interesting feature points. Finally, parameter self- adaptation anti-symmetrical mapping system is formed. The foundation of anti-symmetrical mapping is log-polar transform, because log-polar transform has the characteristic of rotation and scale transform invariability, which make it convenient to match and cognize image. The main work of the paper is as follows:1. Anti-symmetrical mapping model transform: Based on log-polar mapping model, confirm a log-polar mapping model with parameter, analyze the model's characteristic of rotation and scale transform invariability, and how to realize transform and comeback: direct arithmetic and contradictory arithmetic.2. Aiming to scene image matching based on transform image: Primarily analyze 1-D matching projecting transform image to scale and angle axis, 2-D correlation matching and integrative matching scheme. The arithmetic has solved the difficulty brought by scale and rotation change, and has realized the scene matching from 1-D...