-
公开(公告)号：US20190228517A1
公开(公告)日：2019-07-25
申请号：US16370136
申请日：2019-03-29
Inventor： Yuhua CHENG , Chun YIN , Haonan ZHANG , Xuegang HUANG , Ting XUE , Kai CHEN , Yi LI
Abstract: The present invention provides a method for separating out a defect image from a thermogram sequence based on feature extraction and multi-objective optimization, we find that different kinds of TTRs have big differences in some physical quantities, such as the energy, temperature change rate during endothermic process, temperature change rate during endothermic process, average temperature, maximum temperature. The present invention extract these features (physical quantities) and cluster the selected TTRs into L clusters based on their feature vectors, which deeply digs the physical meanings contained in each TTR, makes the clustering more rational, and improves the accuracy of defect separation. Meanwhile, the present invention creates a multi-objective function to select a RTTR for each cluster based on multi-objective optimization. The multi-objective function does not only fully consider the similarities between the RTTR and other TTRs in the same cluster, but also considers the dissimilarities between the RTTR and the TTRs in other clusters, the RTTR is more representative, which guarantees the accuracy of describing the defect outline.
-
公开(公告)号：US20190228221A1
公开(公告)日：2019-07-25
申请号：US16370202
申请日：2019-03-29
Inventor： Chun YIN , Yuhua CHENG , Ting XUE , Xuegang HUANG , Haonan ZHANG , Kai CHEN , Anhua SHI
Abstract: A method for separating out a defect image from a thermogram sequence based on weighted naive Bayesian classifier and dynamic multi-objective optimization. A method extracts these features and classifies the selected TTRs into K categories based on their feature vectors through a weighted naive Bayesian classifier, which deeply digs the physical meanings contained in each TTR, makes the classification of TTRs more rational, and improves the accuracy of defect image's separation. Meanwhile, the multi-objective function does not only fully consider the similarities between the RTTR and other TTRs in the same category, but also considers the dissimilarities between the RTTR and the TTRs in other categories, thus the RTTR selected is more representative, which guarantees the accuracy of describing the defect outline. The initial TTR population approximate solution for multi-objective optimization is chosen according to the previous TTR populations, which makes the multi-objective optimization dynamic and reduces its time consumption.
-
Search Results
2
records
(took 0.017 seconds)
