Synthetic aperture radar (SAR) image fusion focuses to combine relevant information from multiple source images into a single image. Due to the static nature of the environment, and motion in the airborne radar the sensor node receives different source images at different time interval. The captured images at the source may be noisy, incomplete and redundant. Using image fusion techniques a clear and confined image may approximated to the source image can be obtained. In this paper, a Poisson Gaussian noise model is considered that contains Poisson distribution and Gaussian distribution for motion noise and Gaussian noise respectively. An efficient approach for fusion of multiple SAR images based on variance calculation in DCT domain is presented. Due to the simplicity of the proposed method, it can be easily used in real-time applications. The experimental results verify the efficiency improvement of our method both in output quality and complexity reduction in comparison with several conventional techniques.
Consistency verification, discrete cosine transform, image fusion, synthetic aperture radar.