Yarn paper tube classification method with multi-linked fusion optimized template matching

Table 1. Experimental results of each algorithm s
View table
View in Article
Table 1. Experimental results of each algorithm s
图片 SAD SEA 本文算法
心形图案纱纸管 1.372 1.203 0.126
圆形图案纱纸管 1.382 1.215 0.128
笑脸图案纱纸管 1.318 1.209 0.125
心形图案纱纸管(高分辨率) 30.5 22.46 0.920

Table 2. Ablation experiment

View table

View in Article

Table 2. Ablation experiment

匹配算法	匹配位置/pixel	匹配时间/s
算法1：SAD	(0,60)	1.372
算法2：SAD优化算法SEA	(0,60)	1.293
算法3：SEA+自动阈值	(0,60)	0.865
算法4：算法3+十字灰度特征及误差阈值	(0,60)	0.274
算法5：算法4+小波金子塔	(0,60)	0.126

Table 3. Comparison experiment

View table

View in Article

Table 3. Comparison experiment

匹配算法	匹配位置/pixel	匹配时间/s
NCC	(0,60)	2.269
基于动态规划和降采样的 NCC算法^[14]	(2,56)	0.672
自适应步长模板匹配算法^[15]	(0,60)	0.368
SSDA+粗精策略匹配算法^[16]	(0,60)	0.579
本文所提算法	(0,60)	0.126

Tools

Get Citation

Copy Citation Text

Junbo BI, Guoping LI, Meng LI, Haining LIU. Yarn paper tube classification method with multi-linked fusion optimized template matching[J]. Journal of Applied Optics, 2024, 45(2): 365

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites