Bottle Printing Defect Detection Based on CUDA Accelerated Hybrid Image Registration

Liang Gao; Yinghua Liao; Xingyao Si; Qinpeng Luo; Jiajun Zhang; Lang Wang

doi:10.62051/ijcsit.v4n3.14

Authors

Liang Gao
Yinghua Liao
Xingyao Si
Qinpeng Luo
Jiajun Zhang
Lang Wang

DOI:

https://doi.org/10.62051/ijcsit.v4n3.14

Keywords:

Defect detection, Machine vision, Image registration, CUDA parallel acceleration

Abstract

Objective: in order to detect the printing defects of the legal text of aluminum cans and beer bottles, the difference method after image registration based on difference model is used. Aiming at the problems of insufficient accuracy and slow speed of text image registration, a method of accelerating mixed image registration based on CUDA is proposed. Methods: firstly, rough registration based on speeded-up robust features was used to make the detected image basically coincide with the template image; Secondly, the local deformation fine-tuning based on demons elastic registration algorithm is used to improve the registration accuracy; In order to improve the speed of registration, NVIDIA CUDA architecture is used to accelerate and optimize the algorithm in parallel to complete the registration quickly. The improved hybrid registration method can quickly and accurately complete the registration with the template image. After the differential processing, the artifacts are eliminated by threshold segmentation and morphological corrosion expansion processing to complete the defect detection. Results: the experimental results show that the registration accuracy of this method has been significantly improved, and the CUDA acceleration can achieve about 24 times of the acceleration ratio, which greatly shortens the registration time and meets certain detection requirements. Through the detection of 63 defect samples, the success rate is 98.41%. Conclusion: the hybrid registration scheme proposed in this paper can accurately and efficiently complete the bottle label text registration, and realize the defect detection task, which provides an effective solution for the printing defect detection of the legal text of aluminum cans and beer bottles.

Downloads

Download data is not yet available.

References

[1] LI B Y, YANG W H, YANG L, et al. Research on printing defect detection of fertilizer packaging based on template matching [J]. Printing and Digital Media Technology Study, 2023, (02): 39-49. DOI: 10.19370/j.cnki.cn10-1886/ts.2023.02.005

[2] ZHANG D H, ZHU Z F, LI Y Q et al. Research progress of two dimensional image quality defect detection based on machine vision [J]. Packaging Engineering, 2023, 44(23): 198-207. DOI:10.19554/j.cnki.1001-3563.2023.23.024.

[3] WANG X, LIU Z Y, SONG X L et al. Printed Word Defect Detection of Medicinal Glass Bottle Based on the Image Registration [j]. Packaging engineering, 2017, 38 (21): 180-185. DOI:10.19554/j.cnki.1001-3563.2017.21.037.

[4] ZUO C, ZHANG Y B, QI Y S, et al. Detection of surface scratch defects of printing products based on machine vision [j]. Printing and Digital Media Technology Study, 2023, (05): 42-48. DOI:10.19370/j.cnki.cn10-1886/ts.2023.05.004.

[5] HU F S, GUO H, XING J P, et al. Image registration based on label printing defect detection [j]. Optical Technique, 2017, 43 (01): 16-21. DOI:10.13741/j.cnki.11-1879/o4.2017.01.004.

[6] NIU T, LIU L D, WU Y H. Image registration algorithm based on CUDA acceleration [J]. Computer Systems & Applications, 2023, 32(01): 146-155. DOI: 10.15888/j.cnki.csa.008889.

[7] ZHOU L J, XIAO S D, LI S Y, et al. Accelerating digital image processing based on SURF and GPU [J]. Transducer and Microsystem Technologies, 2022, 41(03): 98-100. DOI:10.13873/J.1000-9787(2022)03-0098-03.

[8] GAO S, YUAN X P, GAN S, HU L, BI R, LI R B, LUO W D. UAV image matching method integrating SIFT algorithm and detection model optimization [J]. Spectroscopy and Spectral Analysis, 2022, 42(05):1497-1503.

[9] WU Y Q, WANG Z L. Remote sensing image registration algorithm based on improved surf in wavelet domain [J]. Journal of Tianjin University (Science and Technology), 2017, 50(10): 1084-1092.

[10] GAO Z Z, WEI B, PAN Z K, et al. Image dehazing based on dark channel prior and Hessian regularization [J]. Journal of Graphics, 2020, 41(01):73-80.

[11] ZHANG D, HUANG H, SHANG Z H. Active demons non-rigid image registration algorithm based on mutual information [J]. Laser & Optoelectronics Progress, 2020, 57(16):124-130.

[12] ZHENG W ZHOU Y, LI W H, et al. Research on DTI multi-channel registration based on variable inertia coefficient active demons algorithm [J]. Journal of Hebei University (Natural Science Edition), 2021, 41(04):436-442.

[13] SANTOS-RIBEIOR A, NUTT D J, McGONIGLE J. Inertial demons: a momentum-based diffeomorphic registration framework [C]. Medical Image Computing and Computer-Assisted Intervention-MICCAI 2016: 19th International Conference, Athens, Greece, October 17-21, 2016, Proceedings, Part III 19. Springer International Publishing, 2016: 37-45.