目的 为了体现印刷品不同区域的重要性等级， 提高印刷品缺陷检测的精度， 提出了一种分区域分等级的印刷品缺陷检测方法。方法 根据检测区域的特点和重要性不同， 把印刷品分成不同的区域进行前景提取， 并且设置不同的检测等级； 对印刷品进行符合人眼视觉特性的缺陷识别， 并对提取缺陷进行特征分析。结果 分区域的缺陷检测可以实现不同区域的同时检测， 在检测耗时、 检测准确率及误检率上都优于不分区域的检测方法。结论 基于分区域的印刷品缺陷检测方法能很好地满足印刷品质量检测的需求， 提高印刷品缺陷检测的精度和效率。

To reflect the importance level of the different regions of prints and increase the print defects detection accuracy, a method of print defects detection was proposed based on sub-region and grading. Depending on the characteristics and importance of the detection area, prints were divided into different areas to extracted foreground and different detection levels were set. The print defects were identified according to human vision characteristics, and then the characteristics of extracted were defects analyzed. The method of sub-region defects detection can be implemented at the same time in different areas. It was superior to the method of non-subregion detection in the aspects of detection time consuming, detection accuracy rate and false detection rate. The method of print defects detection based on sub-region can well meet the needs of printing quality detection and improve the accuracy and efficiency of the print defects detection.