【摘要】 由于印刷品质量在线检测过程中存在外界干扰使检测图像产生大量噪声，从而导致检测效率低下和检测精度不高，无法满足工业需求。为解决上述问题，提出了一种基于角点优化的印刷品质量动态检测方法。利用Shi-Tomasi角点算法提取待检图像和模板图像的角点信息以备动态调整时选用，在角点提取过程中根据图像特性淘汰区域冗余角点并利用散列表存储角点信息，避免角点污染并优化算法速度和空间。在动态检测过程中可根据特征相似性调度周围角点信息进行错位判断和坐标调整，完成图像配准。仿真结果表明，所提算法相对于传统的差影检测方法提高了印刷品质量检测精度，满足工业检测需求。更多还原

【Abstract】 Due to the low detection accuracy and slow detection speed of current detection methods, caused by non-ideal factors such as image distortion and position noise, this article represents a dynamic detection method for optimizing the detection of printed images based on corner point optimization. Firstly, image preprocessing and template learning were performed for printed images, and thus to enhance the printed image. Meanwhile, the Shi-Tomasi corner algorithm was used to extract the corners of the printed image and the template image. The point information was used for dynamic adjustment. In the process of corner point extraction, the redundant corner points of the area were eliminated according to the characteristics of the printed image and the corner information was stored in a hash table, avoiding corner point pollution and optimizing the algorithm speed and space. According to the similarity of the corner points, the corner point information around the detected point was scheduled for positioning and coordinates adjustment. In the process of dynamic detection, according to the similarity of the corner points, the corner point information around the target point was scheduled for positioning and coordinates adjustment to complete image matching. The simulation results show that the proposed algorithm improves the speed and accuracy of printing quality detection compared with the traditional difference detection method, and meets the needs of industrial detection.更多还原