Contour Extraction and Quality Inspection for Inner Structure of Deep Hole Components
| Autor: | De Xu, Hu Su, Lei Zhang, Hua-Bin Yang, Zhengtao Zhang, Xinyi Gong |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Basis (linear algebra)
Computer science Machine vision business.industry Detector ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Boundary (topology) 02 engineering and technology Filter (signal processing) 010501 environmental sciences 01 natural sciences Industrial and Manufacturing Engineering Electronic Optical and Magnetic Materials Image (mathematics) Operator (computer programming) Computer Science::Computer Vision and Pattern Recognition Component (UML) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Computer vision Artificial intelligence Electrical and Electronic Engineering business 0105 earth and related environmental sciences |
| Zdroj: | IEEE Transactions on Components, Packaging and Manufacturing Technology. 9:575-585 |
| ISSN: | 2156-3985 2156-3950 |
| Popis: | This paper focuses on the contour extraction for the inner wires of a kind of deep hole component to achieve a high-accuracy inspection. The vision system consisting of a camera and an endoscope is developed to acquire high-quality images of the internal structure. For the acquired images, a contour extraction method is proposed, which could be divided into the following steps. First, the start points of wires referred as the prior information on the component are obtained with a predefined filter, and on the basis, several regions of interest (ROIs) are defined. Second, the multiscale probability of boundary operator is utilized to detect edges in the ROIs. Third, a Brownian motion model is established to calculate the connectivity between edges. The prior information obtained previously is used again to determine the probabilities of the edges belonging to the contours. Finally, the symmetric ratio contour method is used to form the wires’ contours with the edges. In the proposed method, the edges belonging to the wires’ contours are enhanced by making full use of the prior information, resulting in the improvement in accuracy and real-time performance. As evidenced by the experiments, the proposed method can efficiently extract the inner wires’ contours from the component’s image with low-contrast conditions, noises, and shadows. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|