期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2016
卷号：89
期号：1
出版社：Journal of Theoretical and Applied
摘要：Automatic deform detection on automotive body panel is challenging owing to its localization on a large surface, variation in appearance, and their rare occurrences. It is difficult to detect these deforms either by physics-based models or by small-sample statistics using a single threshold. As a result, this problem is focused to derive a set of good-quality deform detection from the surface images. These detection should discriminate the various surface deforms when fed to suitable machine learning algorithms. This research used first order derivative such as gradient filtering and background illumination removal to identify the deform area. An algorithm to segment the deform area has been developed. It localizes deforms employing kernel classifiers, such as Artificial Neural Network (ANN) and Support Vector Machine (SVM). The test result on artificial car door with two types of deformations which is ding and dent
Loading...
