- Research Article
5
- 10.2528/pierb13092608
PASSIVE MILLIMETER WAVE IMAGE DENOISING BASED ON ADAPTIVE MANIFOLDS
- Jan 01, 2014
- Progress In Electromagnetics Research B
- Shujin Zhu + 3 more +3
Since the characters of poor inherent resolution and low signal-to-noise limit the application of the passive millimeter wave (PMMW) image, it is particularly important to improve the resolution and denoise the PMMW image. In this paper, the adaptive manifolds flltering algorithm based on non-local means (AM-NLM) is illustrated in detail. And an improved version of AM-NLM flltering algorithm is proposed for processing the PMMW image. The proposed algorithm flrstly applies the AM-NLM flltering to obtain the basic denoised PMMW image. Then the image enhancement based on Laplacian of Gaussian operator is performed to enhance the edge of the target in PMMW image. Finally, the hard-threshold flltering with difierent thresholds is adopted to fllter each dimension to achieve the flnal flltering response. Experimental results have shown that the proposed PMMW flltering algorithm has better and more satisfactory performance compared to AM-NLM, both in subjective visual efiect and objective image quality metric. Additionally, our proposed algorithm is also available for real PMMW images.
Read more