8 September 2011 An effective recognition algorithm for multiple targets under sea surface background
Author Affiliations +
This paper describes an effective multiple targets recognition algorithm under the circumstances that the targets are with a sea surface background. The proposed algorithm relies on three steps. Firstly, an improved canny operator was adopted to get the edge characteristics of the image templates of airplanes, naval ships and islands , compared to the classical canny operator, the proposed canny operator can generate more accurate edges because of the anisotropic Gaussian filtering and the self-adaptive threshold estimate method it used. Secondly, to calculate the valid objects' positions and generate their edge binary images in the current image, region grown algorithm and improved canny algorithm were adopted. Finally, a hausdorff distance based categorizer was implemented to classify the objects already been located. Experimental results show that the approach performs well in multiple targets recognition in images with sea surface background , the proposed algorithm has strong pertinency and high efficiency to similar problems.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Jian-dong Wang, Jian-dong Wang, Biao Zhang, Biao Zhang, } "An effective recognition algorithm for multiple targets under sea surface background", Proc. SPIE 8193, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications, 81934N (8 September 2011); doi: 10.1117/12.902307; https://doi.org/10.1117/12.902307

Back to Top