Novel adaptive suppression method for SAR image speckles

Abstract

To effectively preserve synthetic aperture radar (SAR) image edges while filtering, an anisotropic diffusion denoising model based on Euler's elastic energy model is presented. First, a minimum-energy variation model is derived from the anisotropic diffusion model. To preserve the image details while filtering, the Euler's elastic energy model is introduced. Then an adaptive step-size iteration scheme is proposed to improve the computational efficiency. The new algorithm using both simulated and real SAR images is validated. The experimental results show that the proposed method not only preserves image details effectively while filtering, but also improves the computational efficiency.

在合成孔徑雷達(synthetic aperture radar,SAR)圖像斑點噪聲抑制處理中,為有效保護圖像細節,提出歐拉彈性能量各向異性擴散去噪模型。該模型將各向異性擴散模型轉化為最小能量變分模型,結合歐拉彈性能量模型的邊界保護和增強能力,在抑制噪聲的同時能更有效地保護和增強細節信息。同時為了提高計算效率,提出自適應變步長去噪算法。仿真和真實SAR圖像的實驗結果表明,該算法不僅在抑制噪聲的同時能夠很好地保護圖像細節,而且有效減少了計算時間、提高了效率。