Design Of The System Based On OMAP For The Underwater Optical Image Matching

Due to water absorption and scattering of the light as well as the complexity of the marine environment and other factors,the underwater light image's quality is low and lack of useful information,which bring many difficulties to identify and locate targets.Faced with the dwindling resources on Earth today,it becomes urgent to exploit the oceans resources and essential to develop the study of underwater light images.The multi-scale theory can analysis characteristics of the object from different angles,and can explain the nature of the object more comprehensively.The article taking the underwater light image as the object of study,designs a underwater light image matching system based on OMAP-L138.The system achieves to extract the image's feature and match the multi-scale image in the Gaussian scale space.The article introduces the multi-scale image matching theory,and introduce an algorithm to extract the feature and match the image in the Gaussian scale space.By the method of extracting extreme point of space in the Gaussian scale space,Characteristics of the image have a good scale,rotation and illumination invariances,and Its noise immunity is strong.Simulation results show the feasibility of the algorithm.Then,The article designs a underwater light image matching system based on OMAP-L138.This system receives real-time images from the PC,and matchs with the reference images which has stored on the SD card.The results are sent to the PC and are displayed on the monitor.Experimental results show that the algorithm for underwater light image matching with good stability?real-time and accuracy.This system realizes the target recognition and tracking through the image matching method.The article,taking the underwater light image as the object of study,carries out a research on the image matching system.The research results may provide a theoretical basis and technological mean to identify and locate underwater targets..