A fast defect detection algorithm for labels based on improved template matching
Labels are widely used in all kinds of tools,glass bottles and other products,marking the relevant important information of the manufacturing. In the production line of manufacturing enterprises,the method of manual inspection labels has become a bottleneck to improve production efficiency.With the acceleration of digital transformation,more and more scholars begin to study the label defect detection algorithm.Taking a glass bottle label as an example,a lightweight tracking algorithm,a positioning algorithm based on pixel mean value,a label detection algorithm based on pixel mean square error,and a label tilt detection and content recognition algorithm based on contour and improved matching process are designed to realize rapid label defect detection,including detecting whether the product is labeled and whether the label is tilted,and identify label content and do the count at the same time.
NETL
NSTL
万方数据
