科研之友
微信
新浪微博
Facebook
分享链接
摘要
In this paper, a fast pre-classified-based super-resolution model has been proposed to over come the problems of degraded imaging in weak-target real-time detection system, specialized to copper defect detection. To accurately characterize the defected image, textural features based on the statistical function of gray-gradient are presented. Furthermore, to improve the effectiveness and practicality of the online detection, a concept of pre-classified learning is introduced and an edge smoothness rule is designed. Some experiments are carried out on defect images in different environments and the experimental results show the efficiency and effectiveness of the algorithm.
- 出版日期2017-7-30
- 单位河海大学
