基于自适应小波的织物疵点自动分割研究

Research on the fabric defect detection based on an adaptive wavelet transform

根据小波正交条件,通过使用疵点图像的疵点区与无疵点区能量均值比为优化目标函数,构造与织物纹理模式相匹配的二维自适应平稳小波.实验证明,构造的小波具有更好的鲁棒性和检测能力.

According to the orthogonal wavelet conditions, a two-dimensional adaptive stationary wavelet which matches the fabric texture is constructed by using the average energy ratio between the defect area and defect-free area as optimal objective function. The experimental result demonstrates that the constructed wavelet in this paper is more robust and has better detection capabilities.