摘要
随着科技的进步,现代化工业生产水平不断提高。而目前生产线中,瓶盖的检测仍靠人力,为实现瓶盖自动化实时检测与筛选,改善采集高速运动瓶盖图像所出现的虚影问题,设计适合高速瓶盖检测的算法。其主要工作包含:分析采集图像中目标边缘数据,创新性地将最小二乘法算法引入到图像处理领域,对目标图像线性化处理,降低虚影程度,从而降低系统对硬件的依赖程度;改进平均阈值分割算法,并结合小面积去除法去除背景杂质点,提取出清晰的目标轮廓;通过仿真技术验证系统算法,得出算法具有简单、快速等特征的结论。通过系统实验,证实该检测系统具有高准确性和高实时性。
With the progress of science and technology, the level of modern industrial production has been continuously raised. At present in the production line, the bottle cap＇ s detection still relies on manpower. To realize automatic real-time detection and screening of bottle caps, and improve the virtual shadow problem of high-speed movement bottle cap images, we design an algorithm suitable for high-speed bottle cap detection. Its main work includes：Analysis of data acquisition object edge image, and introduce the least square method to the field of image processing, the target image is linearized to reduce the virtual shadow degree and reduce the dependence of the system on the hardware; Improved average threshold segmentation algorithm, and combined with a small area removal method to remove the background impurities, to extract a clear target contour; The algorithm is proved by the simulation technology, and it has a simple, fast and other characteristics. Experiments show that the detection system has high accuracy and high real-time performance.
出处
《计算机应用与软件》
2017年第11期223-227,246,共6页
Computer Applications and Software
关键词
机器视觉
最小二乘法
阈值分割
面积滤波
Machine vision Least square method Threshold segmentation Area filtering
