| 摘要: |
| 通过研究经典的缺陷识别算法，提出了一种改进的动态阈值的图像差分算法。该算法采用整体灰度特征与局部相结合的方法，得到差分图像进行印刷缺陷提取。实验运行表明，此方法能快速检出印刷缺陷的位置及面积，准确率高达98％，效率比传统的动态阈值法高26%，在高速生产环境下运行稳定可靠，能够满足印刷品在线质量检测的要求。 |
| 关键词: 印刷品 缺陷提取 动态阈值 |
| DOI： |
| 分类号:TS807 |
| 基金项目:山东省高校科技计划项目（J10LD09） |
|
| A Defect Extraction Algorithm Based on Improved Dynamic Threshold |
|
ZHANG Jin, GAO Jun, FU Yan-wen
|
|
Shandong University of Technology, Zibo 255049, China
|
| Abstract: |
| Through study of classical defect recognition algorithm, an improved dynamic threshold was proposed. The algorithm adopted gray scale characteristics over the global and local combination for image detection. Experimental results showed that the algorithm can fast detect all types of defect location and size with accuracy of 98%, and efficiency is 26% higher than traditional threshold algorithm. The algorithm is stable and reliable operation in high speed production environment and can meet requirements of online quality detection. |
| Key words: printed matter defect extraction dynamic threshold |
