Study Of Extraction Of Bridge And Water Front From Inland River Maritime Radar Image Based On Image Processing

Radar as an active, real-time access to technology of ship navigation channelstate information and environmental information plays an important part of maritimesafety and supervision. Through the automatic extraction of coastline radar imagesof the bridge, can provide a technical basis for high precision radar image and thesuperposition of electronic map, contribute to a more accurate extraction of the realship sailing area, improving radar plotting ability, strengthen maritime securitycapacity.This paper analyzes the basic characteristics of maritime radar images of theinland radar and generated image noise, through a variety of image processingmeans, realize the extraction of coastline in radar images of the bridge, providetechnical support for automatic plotting superposition electronic chart display andradar. In this paper, the main research work is as follows:(1) According to the characteristics of radar images, carried out the study on themethod of image pretreatment. Get the radar image background generation throughinter frame difference method, extract the static target from the inland maritimeradar images; comparison and analysis of the advantages and disadvantages andapplicability of various filtering algorithm, realization of noise suppression based onmedian filter; analysis law of non target Unicom domain pixel statistics, on the basisof the extracted shoreline and bridge the Unicom domain; comparative analysis ofthe treatment effect of different edge operator, and implemented for the shorelinebridge image edge extraction processing;(2) Carried out a study on the extraction method of bridge based on thecharacteristics of target shoreline angle. Using the improved Hough transform toacquire the shoreline of bridge target inclination angle information, and the Houghtransform the obtained angle information into Gabor wavelet, extracted from theedge image of shoreline bridge target in current; (3) For radar image environment under different waters, method of extractingtarget shoreline bridge test proposed and verified the validity of the method ofextraction.Through this study, to realize the automatic extraction of coastline in radarimage bridge target, after the actual test of various types of radar images underdifferent environments, we can see that the proposed algorithm has good robustnessand lay a technical foundation for the later research of high precision radar imageand the superposition of electronic map, also contributed to the water traffic safetyguarantee ability promotion.