Defect Detection for Glass Seal Insulated Terminals Based on Sector Neighborhood Feature Engineering
-
摘要: 该文提出一种基于特征工程的玻璃封装绝缘端子外观质量检测算法以替代人工肉眼检测。首先，基于绝缘端子的形状先验将待检测区域划分为若干个扇区。其次，根据玻璃封装绝缘端子图像特点提出扇区基本特征数据、扇区灰度变化率、扇区反光特征和扇区方向统计特征等4大类扇区特征提取方法，采用梯度提升决策树(GBDT)进行粗分类。为了更全面地表征扇区特性，基于粗分类结果融合最近邻扇区提出扇区近邻(SN)特征提取方法。最后，将扇区近邻特征和扇区特征都输入到GBDT分类器进行精细分类，获得检测结果。实验结果表明，提出的特征工程方法能够在合理检测时间内取得较好的检测性能，交并比为97.45%，F1为0.987，优于现有类似检测方法。Abstract: In this paper, an appearance quality inspection method for glass seal insulated terminals is proposed based on feature engineering to replace the current manual inspection. First, the inspecting region in the glass seal insulated terminal image is divided into many sectors based on the shape prior of the terminal. Second, considering the characteristics of the image, four categories of sector features are designed, such as sector basic features, sector gray change rates, sector reflection features and sector direction statistical features. Then, they are input into a Gradient Boosting Decision Tree (GBDT) for rough classification. Next, to characterize excellently the sectors, a new image feature called Sector Neighborhood (SN) feature is designed by combining sector features of the nearest neighbor sectors and rough classification results for these sectors. Finally, the SN features and sector features are input into the GBDT for fine classification, which indicates final quality inspection. The experimental results indicate that the proposed method can achieve better inspection performance with reasonable inspection time compared to the existing inspection methods, which has 97.45% IoU and 0.987 F1.
-
[1] GUO Hongwei, DANG Mengyang, LIU Lei, et al. Alkali barium glasses for hermetic compression seals: Compositional effect, processing, and sealing performance[J]. Ceramics International, 2019, 45(17): 22589–22595. doi: 10.1016/j.ceramint.2019.07.290 [2] SHEN Ziqin, ZHANG Yong, CHEN Yongzhou, et al. Effect of pre-oxidization condition on glass-to-metal sealing[J]. Journal of Non-Crystalline Solids, 2019, 521: 119488. doi: 10.1016/j.jnoncrysol.2019.119488 [3] THOMPSON L M, MAUGHAN M R, RINK K K, et al. Thermal induced stresses in bridge-wire initiator glass-to-metal seals[J]. Journal of Electronic Packaging, 2007, 129(3): 300–306. doi: 10.1115/1.2753920 [4] LEI Dongqiang, WANG Zhifeng, and LI Jian. The calculation and analysis of glass-to-metal sealing stress in solar absorber tube[J]. Renewable Energy, 2010, 35(2): 405–411. doi: 10.1016/j.renene.2009.05.021 [5] FAN Zhichun, HU Kangjia, HUANG Zhiyong, et al. Optimized sealing process and real-time monitoring of glass-to-metal seal structures[J]. Journal of Visualized Experiments, 2019(151): e60064. doi: 10.3791/60064 [6] WANG Yang, HE Wei, MO Yunqi, et al. Failure analysis on the chip capacitor[C]. 2009 IEEE Circuits and Systems International Conference on Testing and Diagnosis, Chengdu, China, 2009. [7] KRIEGER V, WONDRAK W, DEHBI A, et al. Defect detection in multilayer ceramic capacitors[J]. Microelectronics Reliability, 2006, 46(9/11): 1926–1931. doi: 10.1016/j.microrel.2006.07.082 [8] LIU Qunpo, WANG Manli, WANG Gaowei, et al. Detection algorithm of porosity defect on surface of micro-precision glass encapsulated electrical connectors[J]. Journal of Robotics, Networking and Artificial Life, 2020, 7(3): 212–216. doi: 10.2991/jrnal.k.200909.015 [9] LIU Qunpo, WANG Mengke, LIU Zonghui, et al. Defect detection of micro-precision glass insulated terminals[J]. Journal of Robotics, Networking and Artificial Life, 2021, 8(1): 18–23. doi: 10.2991/jrnal.k.210521.005 [10] LIU Qunpo, WANG Mengke, and HANAJIMA N. Defect sample generation system based on DCGAN for glass package electrical connectors[C]. Proceedings of 2020 Chinese Intelligent Systems Conference, Shenzhen, China, 2020: 434–441. [11] FU Li, ZHANG Shuai, GONG Yu, et al. Medicine glass bottle defect detection based on machine vision[C]. 2019 Chinese Control and Decision Conference (CCDC), Nanchang, China, 2019. [12] LIN H D and CHIU S W. Flaw detection of domed surfaces in LED packages by machine vision system[J]. Expert Systems with Applications, 2011, 38(12): 15208–15216. doi: 10.1016/j.eswa.2011.05.080 [13] ZHOU Xianen, WANG Yaonan, ZHU Qing, et al. A surface defect detection framework for glass bottle bottom using visual attention model and wavelet transform[J]. IEEE Transactions on Industrial Informatics, 2020, 16(4): 2189–2201. doi: 10.1109/TII.2019.2935153 [14] FRIEDMAN J H. Greedy function approximation: A gradient boosting machine[J]. The Annals of Statistics, 2001, 29(5): 1189–1232. doi: 10.1214/aos/1013203451 -
下载:
图(4) / 表(1)
计量
- 文章访问数: 384
- HTML全文浏览量: 219
- PDF下载量: 64
- 被引次数: 0
下载:
