Algorithm for Fast Registration of Radar Images

Defence Science Journal

Dipti Deodhare¹  Subrata Rakshit¹ 

[1] Centre for Artificial Intelligence & Robotics, Bangalore

关键词: Image processing;    radar imagery;    battlefield surveillance;    image registration;    image matching;    algorithms;    radar images;    guidance;   

DOI  :  

学科分类：社会科学、人文和艺术（综合）

来源: Defence Scientific Information & Documentation Centre

【 摘 要 】

Radar imagery provides an all-weather and24 h coverage, making it ideal for critical defence applications. Insome applications, multiple images acquired of an area need to beregistered for further processing. Such situations arise for battlefieldsurveillance based on satellite imagery. The registration has to bedone between an earlier (reference) image and a new (live) image. Forautomated surveillance, registration is a prerequisite for changedetection. Speed is essential due to large volumes of data involved andthe need for quick responses. The registration transformation is quitesimple, being mainly a global translation. (Scale and rotationcorrections can be applied based on known camera parameters). Thechallenge lies in the fact that the radar images are not as feature-richas optical images and the image content variation can be as high as 90per cent. Even though the change on the ground may not be drastic,seasonal variations can significantly alter the radar signatures ofground, vegetation, and water bodies. This necessitates a novel approachdifferent from the techniques developed for optical images. Analgorithm has been developed that leads to fast registration of radarimages, even in the presence of specular noise and significant scenecontent variation. The key features of this approach are adaptability tosensor/terrain types, ability to handle large content variations andfalse positive rejection. The present work shows that this algorithmallows for various cost-performance trade-offs, making it suitable for awide variety of applications. The algorithm, in variouscost-performance configurations, is tested on a set of ERS images.Results of such tests have been reported, indicating the performance ofthe algorithm for various cost-performance trade-offs.

【 授权许可】

Unknown