基于改进自适应阈值的织物疵点检测算法研究

Research on fabric defect detection algorithm based on improved adaptive threshold

提出了一种改进的自适应阈值算法。首先引入前置低通滤波器,降低局部纹理的复杂性,同时较好地保持疵点区域,然后对滤波后的织物图像进行自适应阈值分割,可获得较好的分割效果。实验结果表明,该算法对简单纹理和复杂纹理的织物图像均具有较好的分割结果。

In this paper,we propose an improved adaptive threshold algorithm based on traditional adaptive threshold algorithm. By introducing pre-fihering to reduce local textural complexity and keep the defect region well, we can process the filtered image using adaptive threshold algorithm and get a better segment result.The experiment results demonstrate that this algorithm can get a better detection result to both simple and complex textural fabric image.