2021
DOI: 10.3390/app11104402
|View full text |Cite
|
Sign up to set email alerts
|

Inspection System for Vehicle Headlight Defects Based on Convolutional Neural Network

Abstract: This paper proposes a method to detect the defects in the region of interest (ROI) based on a convolutional neural network (CNN) after alignment (position and rotation calibration) of a manufacturer’s headlights to determine whether the vehicle headlights are defective. The results were compared with an existing method for distinguishing defects among the previously proposed methods. One hundred original headlight images were acquired for each of the two vehicle types for the purpose of this experiment, and 20… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
2
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
2
1

Relationship

1
2

Authors

Journals

citations
Cited by 3 publications
(2 citation statements)
references
References 12 publications
0
2
0
Order By: Relevance
“…in the glass on the front of smartphones. As preprocessing, flood filling ( Figure 18 c) and image rotation ( Figure 18 d) were applied [ 6 ], after iterative binarization ( Figure 18 b) of the grayscale image ( Figure 18 a). For an objective performance analysis, the performance of the proposed method was compared with those of the line-based [ 7 , 14 ] and corner-based reference point extraction methods [ 8 , 9 , 14 ], among the existing reference point search algorithms [ 8 , 9 , 14 ].…”
Section: Experiments and Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…in the glass on the front of smartphones. As preprocessing, flood filling ( Figure 18 c) and image rotation ( Figure 18 d) were applied [ 6 ], after iterative binarization ( Figure 18 b) of the grayscale image ( Figure 18 a). For an objective performance analysis, the performance of the proposed method was compared with those of the line-based [ 7 , 14 ] and corner-based reference point extraction methods [ 8 , 9 , 14 ], among the existing reference point search algorithms [ 8 , 9 , 14 ].…”
Section: Experiments and Resultsmentioning
confidence: 99%
“…The feature extraction methods for alignment correction involve the use of a line [ 1 , 2 ], integral histograms [ 3 , 4 ], projection-based integral histograms [ 5 ], the long axis [ 6 ], Hough Line-based reference feature extraction [ 7 ], Harris Corner-based feature extraction [ 8 ], and Moravec Corner-based feature extraction [ 9 ]. The method involving the use of a line, which is a feature included in the part, has the advantage of rapidly correcting the alignment of products with line components; however, it is difficult to apply to round-shaped parts.…”
Section: Introductionmentioning
confidence: 99%