| 摘要: |
| 实孔径毫米波雷达图像乘性噪声使得图像目标难以识别。为了降低乘性噪声水平，提高道路环境识别的准确性，结合图像分割技术，提出了一种基于全变分同态滤波的雷达图像噪声抑制算法。该算法首先对图像进行分割以提取目标的边缘信息和图像的平滑区域,随后在对数域中利用同态滤波有效抑制平滑区域中的乘性噪声,最后通过全变分方法对图像的各个区域进行精细化去噪处理。实验结果表明，所提方法在噪声抑制图像平滑方面，其等效视数接近无穷大；在边缘保持客观指标衡量方面，其β指数值也优于参考方法。 |
| 关键词: 毫米波雷达图像 图像去噪 全变分 同态滤波 |
| DOI：10.20079/j.issn.1001-893x.240604005 |
|
| 基金项目:江苏省自然科学基金面上项目（BK20231336）；南通市社会民生科技计划-重点专项(MS12022005) |
|
| Noise Reduction of Real Aperture Millimeter-wave Radar Images Using Total Variation Homomorphic Filtering |
| HUANG Tonghui,XU Zhihuo |
| (School of Transportation and Civil Engineering,Nantong University,Nantong 226019,China) |
| Abstract: |
| The multiplicative noise in real aperture millimeter-wave radar images makes it difficult to identify targets.To reduce the level of multiplicative noise and improve the accuracy of road environment recognition,by incorporating image segmentation techniques,an innovative radar image noise suppression algorithm based on total variation homomorphic filtering is proposed.The algorithm first segments the image to extract edge information and smooth regions.Then,in the logarithmic domain,homomorphic filtering is applied to effectively suppress multiplicative noise in the smooth regions.Finally,total variation is used for refined denoising across different regions of the image.Experimental results show that the proposed method achieves an equivalent number of looks approaching infinity in terms of noise suppression and image smoothing.In terms of edge preservation,the β-index of the proposed method also outperforms that of the reference methods. |
| Key words: millimeter-wave radar images image denoising total variation homomorphic filtering |
