TY - GEN
T1 - Defect detection based on edge detection and connected region filtering algorithm
AU - Pan, Yiwei
AU - Lu, Yingcong
AU - Dong, Shuhang
AU - Zhao, Zihan
AU - Zhao, Zhanfeng
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/7
Y1 - 2019/7
N2 - In the production process of the print heads, due to the performance of the device and other external factors, some dead pixels affecting the quality of the print heads are formed on the surfaces of them. With the development of image processing technology, it is a good choice to use the related method of image processing to detect dead pixels. Aiming at the problem of accurately detecting the dead pixels of the print head, this paper proposes an identification method using the edge detection algorithm and the connected region filtering algorithm based on linear fitting. Firstly, this paper uses the Retinex algorithm to enhance the original image, and then, minimum filtering and wiener filtering are adopted in order to make further enhancement and remove noise. Secondly, edge detection algorithm is adopted to detect the textures of interference areas and dead pixels when identifying them. Additionally, this paper proposes the algorithm of screening connected regions based on linear fitting to eliminate all interference areas. Finally, the author reserves the dead pixels that need to be identified and then positions them. The experimental results show that this method can accurately detect the dead pixels in the line CCD scanned image and can be applied to at least five types of print head surface dead pixels.
AB - In the production process of the print heads, due to the performance of the device and other external factors, some dead pixels affecting the quality of the print heads are formed on the surfaces of them. With the development of image processing technology, it is a good choice to use the related method of image processing to detect dead pixels. Aiming at the problem of accurately detecting the dead pixels of the print head, this paper proposes an identification method using the edge detection algorithm and the connected region filtering algorithm based on linear fitting. Firstly, this paper uses the Retinex algorithm to enhance the original image, and then, minimum filtering and wiener filtering are adopted in order to make further enhancement and remove noise. Secondly, edge detection algorithm is adopted to detect the textures of interference areas and dead pixels when identifying them. Additionally, this paper proposes the algorithm of screening connected regions based on linear fitting to eliminate all interference areas. Finally, the author reserves the dead pixels that need to be identified and then positions them. The experimental results show that this method can accurately detect the dead pixels in the line CCD scanned image and can be applied to at least five types of print head surface dead pixels.
KW - Algorithm of screening connected regions
KW - Dead pixel detection of print head
KW - Edge detection
KW - Single-scale Retinex algorithm
UR - https://www.scopus.com/pages/publications/85071948642
U2 - 10.1109/CISCE.2019.00065
DO - 10.1109/CISCE.2019.00065
M3 - 会议稿件
AN - SCOPUS:85071948642
T3 - Proceedings - 2019 International Conference on Communications, Information System, and Computer Engineering, CISCE 2019
SP - 254
EP - 259
BT - Proceedings - 2019 International Conference on Communications, Information System, and Computer Engineering, CISCE 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 International Conference on Communications, Information System, and Computer Engineering, CISCE 2019
Y2 - 5 July 2019 through 7 July 2019
ER -