3D reconstruction of cigarette end face topography and study of key quality defects

Aiming at the problems of difficult high-precision quantification of cigarette end face defects and lack of systematic research on key influencing parameters for end face tobacco dropping, this paper constructs a high-precision detection system based on line structured light, achieving three-dimensional reconstruction of cigarette end face topography. Based on the reconstructed point cloud data, quantitative models for typical defects including loose ends, end face crush, exposed cut tobacco, and end face flatness were established, and key characterization parameters were defined, including maximum loose ends depth, loose ends area ratio, crush depth, crush arc ratio, exposed cut tobacco length, and end face flatness. The system repeatability test results show that the RSD of each parameter is below 1.5%. Spearman rank correlation analysis indicates that end face flatness, maximum loose ends depth, and loose ends area ratio show extremely significant positive correlations with content of loss of cut tobacco (P-value < 0.001), while Crush arc ratio shows a significant negative correlation (P-value < 0.01), and crush depth exhibits no significant correlation with tobacco dropping amount. The research achieved quantitative characterization of multiple defects on cigarette end faces and identified the key parameters of content of loss of cut tobacco, thereby laying a theoretical foundation for intelligent detection and quantitative evaluation of cigarette end face quality.