A Novel Method for Detecting the Width of Tobacco Filaments Based on Coordinate Point Fitting
Jinxin Tie, Yue-Feng Wang, Weizheng Zhang, Meng Li, Daojian Yang, Jinhua Song, LI Mao-song, Junhui Wang, Mingfang Zheng, Chunbo Yuan, Yue Long, Yujie Wu, Jie Li · 2023
Abstract. The width of the tobacco filament (WTF) is a significant process parameter in cigarette production as it has a crucial impact on both the physical quality and sensory experience of the product. The detection methods currently employed for WTF mainly involve manual visual inspection or low-frequency offline inspection, which lack representation, accuracy, and efficiency. To enhance the detection efficiency and accuracy, a method for detecting WTF images utilizing the coordinate point fitting method (CPFM) is proposed. We compare the performance of CPFM with the conventional Hough transform method (HTM) and the circle with different radii method (CDRM). Based on experimental results, CPFM has a relative error of 5.3%, an absolute error of 0.085 mm, and a maximum relative standard deviation of 1% in six repetitive experiments. This suggests that the suggested CPFM technique is dependable and dependable and fulfills the statistical demands of metrology. By implementing the WTF image detection method, which relies on the coordinate point fitting methodology, detection efficiency and accuracy are substantially increased. The technique necessitates rapid and exact WTF detection and surpasses prior approaches significantly. Compared to traditional methods, the CPFM method exhibits distinct benefits in detecting WTF. It is probable that the method will become widely used in cigarette production to improve both product quality and production efficiency.