论文部分内容阅读
为了解决SAR图像受相干斑噪声干扰和震后发生形变而识别率偏低的问题,提出了一种新的仿射、形变不变特征-热核特征,并将该特征用于SAR图像目标识别。首先采用推广的核模糊C-均值方法分割SAR图像,提取SAR图像目标形状;接着对目标形状进行Delaunay三角剖分,采用余切权重法对Laplace-Beltrami Operator离散化,通过离散化Laplace-Beltrami Operator特征值、特征向量求每一点热核特征;然后采用谱距离公式对点点间热核距离计算,转化为距离分布表示目标形状的热核特征;最后采用L1相似性准则对图像进行相似性度量,得到识别结果。实验表明:与经典的Hu不变矩方法相比,对于仿射变换和发生形变的SAR图像,该方法都具有更高的识别率。因此,基于热核特征的SAR图像识别方法是一种更加有效的识别方法。
In order to solve the problem of low recognition rate of SAR images affected by speckle noises and post-earthquake deformation, a new affine and deformation invariant feature called thermonuclear feature is proposed and used to target SAR image recognition . First, the SAR image is segmented using the extended kernel-fuzzy C-means method to extract the target shape of the SAR image. Then the target shape is Delaunay triangulated, the Laplace-Beltrami Operator is discretized using the cotangent weight method, and the Laplace-Beltrami Operator Eigenvalue and eigenvector, and then calculate the thermonuclear distance of the target shape by the spectral distance formula. Then, the similarity of the image is calculated by L1 similarity criterion, Get recognition result. Experiments show that compared with the classical Hu moment invariants, this method has a higher recognition rate for affine transformation and deformation of SAR images. Therefore, the method based on thermonuclear feature of SAR image recognition is a more effective method of identification.