人造草坪生产线簇绒机漏纱实时检测方法
|
摘要
针对人工检测簇绒机漏纱时劳动强度大、漏纱检测不及时导致草坪质量差、漏纱修补工作量大的问题,提出了一种基于机器视视觉的漏纱实时检测方法。在检测到漏纱时,该方法使簇绒机自动停机。研究了YUV、HIS等颜色模型的原理,选择合理模型对ROI分色。基于灭点和透视原理建立畸变校正模型,使检测到的纱线姿态统一;根据纱线分布,设计了基于图像灰度值纵向投影加和的纱线位置及漏纱实时检测算法。该算法采用固定时间间隔内,纱线所在区域灰度值加和对比方法检测漏纱,对光照变化不敏感、鲁棒性强,可较好地应用于受自然光影响的人造草坪生产线。检测结果表明,该设计能够及时有效地在簇绒机发生漏纱时使簇绒机自动停机,减小了工人的工作量,提高了生产效率。
Aiming at the problems of manual detection of the yarn missing of the tufting machine in artificial lawn production,such as high labor intensity,poor lawn quality due to inadequate detection of yarn missing and heavy workload in patching up,a real-time detection method of yarn missing based on machine vision is proposed to realize automatic shutdown of tufting machine when yarn missing is detected. The principle of color models such as YUV and HIS is studied,and reasonable model is selected for ROI color separation. A distortion correction model is established based on vanishing point and perspective principle to unify the detected yarn attitudes. According to the yarn distribution,a real-time yarn location and yarnmissing detection algorithm based on image gray value longitudinal projection summation is designed. The algorithm detects the yarn missing by gray value summation and contrast method in fixed time interval,which is insensitive to light change and has strong robustness,and it can be better used in artificial lawn production line affected by natural light. The results show that the design can stop the tufting machine automatically when the yarn missing occurs in time and effectively,reduce the workload of workers and improve the production efficiency.
Aiming at the problems of manual detection of the yarn missing of the tufting machine in artificial lawn production,such as high labor intensity,poor lawn quality due to inadequate detection of yarn missing and heavy workload in patching up,a real-time detection method of yarn missing based on machine vision is proposed to realize automatic shutdown of tufting machine when yarn missing is detected. The principle of color models such as YUV and HIS is studied,and reasonable model is selected for ROI color separation. A distortion correction model is established based on vanishing point and perspective principle to unify the detected yarn attitudes. According to the yarn distribution,a real-time yarn location and yarnmissing detection algorithm based on image gray value longitudinal projection summation is designed. The algorithm detects the yarn missing by gray value summation and contrast method in fixed time interval,which is insensitive to light change and has strong robustness,and it can be better used in artificial lawn production line affected by natural light. The results show that the design can stop the tufting machine automatically when the yarn missing occurs in time and effectively,reduce the workload of workers and improve the production efficiency.
引文
[1]芦长椿.人造草坪的技术发展及应用[J].纺织导报,2011(6):63-70.
[2]厉荣宣,史东东.工件表面裂纹机器视觉检测研究[J].自动化仪表,2017,38(9):85-94.
[3]虎晓红,李炳军,刘芳.多颜色空间中玉米叶部病害图像图论分割方法[J].农业机械学报,2013,44(2):177-181.
[4] MITTAL A,SOFAT S,HANCOCK E. Anefficient scheme for color edge detection in uniform color space[J]. Autonomous and Intelligent Systems,2012,26(37):260-267.
[5]马玲,张晓辉. HSV颜色空间的饱和度与明度关系模型[J].计算机辅助设计与图形学学报,2014,26(8):1272-1278.
[6] AHIRWAL B,KHADTARE M,RAKESH M. FPGA based system for color space transformation RGB to YIQ and YCbCr[J]. IEEE Computer Society,2007,32(9):1345-1349.
[7] Hovhannisyan,A. Chilingariana. median filtering algorithms for multichannel detectors[J]. Advances in Space Research. 2011,47(9):1544-1557.
[8] SANDRA S. Application of genetic algorithm in median filtering[J].Technical University of Wroclaw. 2008,34(4),79-85.
[9]蔡国榕,李绍滋,吴云东,等.一种透视不变的图像匹配算法[J].自动化学报,2013,39(7):1053-1061.
[10]杨永跃,邓善熙,王训四.视觉成像畸变的直线透视校正[J].应用科学学报,2006(6):627-632.
[11]周葳.二维视频中静态场景深度图获取的算法分析[D].哈尔滨:哈尔滨工业大学,2012.
[12]斯蒂格,尤里奇,威德曼.机器视觉算法与应用[M].北京:清华大学出版社,2008.
[2]厉荣宣,史东东.工件表面裂纹机器视觉检测研究[J].自动化仪表,2017,38(9):85-94.
[3]虎晓红,李炳军,刘芳.多颜色空间中玉米叶部病害图像图论分割方法[J].农业机械学报,2013,44(2):177-181.
[4] MITTAL A,SOFAT S,HANCOCK E. Anefficient scheme for color edge detection in uniform color space[J]. Autonomous and Intelligent Systems,2012,26(37):260-267.
[5]马玲,张晓辉. HSV颜色空间的饱和度与明度关系模型[J].计算机辅助设计与图形学学报,2014,26(8):1272-1278.
[6] AHIRWAL B,KHADTARE M,RAKESH M. FPGA based system for color space transformation RGB to YIQ and YCbCr[J]. IEEE Computer Society,2007,32(9):1345-1349.
[7] Hovhannisyan,A. Chilingariana. median filtering algorithms for multichannel detectors[J]. Advances in Space Research. 2011,47(9):1544-1557.
[8] SANDRA S. Application of genetic algorithm in median filtering[J].Technical University of Wroclaw. 2008,34(4),79-85.
[9]蔡国榕,李绍滋,吴云东,等.一种透视不变的图像匹配算法[J].自动化学报,2013,39(7):1053-1061.
[10]杨永跃,邓善熙,王训四.视觉成像畸变的直线透视校正[J].应用科学学报,2006(6):627-632.
[11]周葳.二维视频中静态场景深度图获取的算法分析[D].哈尔滨:哈尔滨工业大学,2012.
[12]斯蒂格,尤里奇,威德曼.机器视觉算法与应用[M].北京:清华大学出版社,2008.