Abstract—To detect and recognize the small targets under complex background, a new segmentation algorithm based on wavelet transform and mathematical morphology is presented. Firstly, the original image is decomposed in three levels by wavelet transform and the high-frequent images are fused by minimum mean square error algorithm to suppress the background and enhance the target. Secondly, the fused images are segmented with an adaptive threshold to get the centroid coordinates of the target area. Thirdly, the original images are segmented accurately with region growing algorithm to get the precise contours and centroid coordinates of the target area. The results of flight test show that the segmentation algorithm has high precision and accuracy, as well as strong engineering adaptability.

Index Terms—WAVELET Transform, Mathematics Morphologic, small target, image segmentation

Cite: Yanghua Li, Jie Zhao, and Bo Mo, "Small Target Segmentation Algorithm Based on Wavelet Transform and Mathematics Morphologic," International Journal of Electronics and Electrical Engineering, Vol. 1, No. 2, pp. 76-79, June 2013. doi: 10.12720/ijeee.1.2.76-79