Object extraction method with accurate contour information

The invention relates to an object extraction method with accurate contour information. The method mainly includes the following steps: step 1, color space conversion; step 2, component vectorization;step 3, principal component analysis; step 4, matrix reconstruction and normalization; step 5, morphological closure operations; step 6, inverse operations; step 7, two-dimensional single-scale wavelet transform; step 8, automatic threshold calculation and merging; step 9, binarization processing and edge extraction; and step 10, post-processing. The contour information of an object can be accurately extracted, and the method can be applied to various target recognition fields of lesion information extraction of medical images, shape recognition of industrial production lines and the like. 本发明涉及种具有准确轮廓信息的对象提取方法。主要包括如下步骤：步骤1，颜色空间转换；步骤2，分量向量化；步骤3，主成分分析；步骤4，矩阵重构及归化；步骤5，形态学闭运算；步骤6，反运算；步骤7，二维单尺度小波变换；步骤8，自动阈值计算及合并；步骤9，二值化处理及边缘提取；步骤10，后处理。能够准确提取对象的轮廓信息，可以应用于医学图像的病变信息提取、工业生产线的形状识别等多种目标识别领域。