|
|
Tracking algorithm of visible particles in transparent
liquid pharmaceutical |
XU Xue-mei, LI Li-xian, ZHANG Jian-yang, NI Lan, HUANG Zheng-yu, CAO Jian |
Department of Physics and Electronics, Central South University, Changsha 410083, China |
|
|
Abstract The tracking algorithm of visible particles in transparent liquid pharmaceutical was analyzed based on intelligent vision technology in order to realize the automatic quality testing in the production of liquid pharmaceutical and improve the speed and accuracy of quality inspection. The median filter was adopted to denoise the corresponding images. The second difference with the energy accumulation algorithm was used to detect particles. Then the flood fill method and real-time motion template were applied respectively to extract the morphological characteristics and the direction of movement about particles. The region-based tracking model was established based on the analysis of particles’ moving tracking. The selection principle of size and location of the search window was given. Different tracking strategy was adopted according to the differences of speed and direction in order to solve the tracking difficulties when particles disappear or overlap. Results show that the wrongtracking rate and the missingtracking rate of the improved regional matching algorithm is 0% and 2.5% respectively, and the tracking accuracy can reach 97.5%. The improved algorithm has a better real-time quality and the needs of real-time production can be satisfied.
|
Published: 01 October 2012
|
|
透明液体药剂中可见异物跟踪算法
为了实现液体药剂生产过程中的自动化质量检测,提高质检的速度和准确性,在智能视觉的基础上对透明液体药剂中可见异物的跟踪算法进行研究.采用中值滤波器对输液图像进行去噪,应用二次差分与能量累积算法对液体药剂中的可见异物进行检测,分别运用漫水填充法和实时运动模板提取异物目标的形态特征和运动特征;在分析异物的运动轨迹基础上,建立基于运动特征的区域匹配跟踪模型,给出搜索窗口的定位与尺寸大小选择原则,根据速度大小和运动方向的差异采取不同的跟踪策略,解决了异物目标在消失和重叠情况下的跟踪难题.实验结果表明,改进的区域匹配跟踪算法的错跟率和漏跟率分别为0%和2.5%,跟踪的正确率可以达到97.5%,且具有较高的实时性,能够满足实际生产的需要.
|
|
[1] 国家药典委员会.中华人民共和国药典[M].2版.北京:化学工业出版社,2005: 311.
[2] MALAMAS E N,PETRAKIS E G M, ZERVAKIS M, et al.A survey on industrial vision systems,applications and tools,Image and Vision Computing 21 [J].Image and Vision Computing,2003,21(2): 171-188.
[3] CHEEZUM M K,WALKER W F,GUILFORD W H,et al. Quantitative comparison of algorithms for tracking single fluorescent particles [J].Biophysical Journal,2001,81(10): 2378-2388.
[4] BOYER K L,OZGUNER T.Robust online detection of pipeline corrosion from range data [J].Machine Vision and Applications,2001,12(6): 291-304.
[5] SHAFAIT F,IMRAN S M,KLETTEMATZAT S.Fault detection and localization in empty warter bottles through machine vision [C]∥Emerging Technology Conference. San Diego: \ [s. n.\], 2004: 30-34.
[6] 李杨果,王耀南,王威.基于机器视觉的大输液智能灯检机研究[J].光电工程,2006,33(11): 69-74.
LI Yangguo,WANG Yaonan,WANG Wei.Intelligent transfusion liquor inspector based on machine vision [J].OptoElectronic Engineering,2006,33(11): 69-74.
[7] 张辉,王耀南,周博文.基于机器视觉的液体药品异物检测系统研究[J].仪器仪表学报,2009,30(3): 548-553.
ZHANG Hui, WANG Yaonan, ZHOU Bowen. Research on foreign substance detection system for medicinal solution based on machine vision [J]. Chinese Journal of Scientific Instrument, 2009, 30(3): 548-553.
[8] 周博文,王耀南,葛继.基于机器视觉的医药注射剂智能检测系统研究[J].机器人,2009,31(1): 53-60.
ZHOU Bowen, WANG Yaonan, GE Ji. A machinevisionbased intelligent inspection system for pharmaceutical injections [J].Robot,2009,31(1): 53-60.
[9] 何成,王耀南.灌装液体药品质量的机器视觉检测与识别[J].中南大学学报,2009,40 (4): 1003-1007.
HE Cheng,WANG Yaonan.Machine vision detection and recognition of quality of filling liquid medicine [J].Journal of Central South University,2009,40(4): 1003-1007.
[10] 鲁娟,王耀南,余洪山.大输液中可见异物智能在线检测系统设计[J].计算机测量与控制,2008,16 (12): 1802-1805.
LU Juan,WANG Yaonan,YU Hongshan.Research on online detection system for foreign substances in a transfusion bottle of medicinal solution [J].Computer Measurement and Control,2008,16(12): 1802-1805.
[11] 杨福刚,孙同景,宋松林.基于人工免疫算法的药液颗粒异物检测方法[J].电子测量与仪器学报,2008,22(1): 20-24.
YANG Fugang,SUN Tongjing,SONG Songlin.Artificial immune algorithm based particle detection method for liquid in pharmaceutical container [J].Journal of Electronic Measurement and Instrument,2008, 22(1): 20-24.
[12] YEN J C,CHANG F J,CHANG S.A new criterion for automatic multilevel thresholding [J]. IEEE Transactions on Image Processing,1995,4(3): 370-378.
[13] MINGKUEI H.Visual pattern recognition by moment invariants [J].IRE Transaction on Information Theory,1962,8(2): 179-187.
[14] GARY R B,JAMES W D. Motion segmentation and pose recognition with motion history gradients [J].Machine Vision and Applications, 2002(13): 174-184.
[15] DUNCAN DY P,MINH N D.Directional multiscale modeling of images using the contourlet transform [J].IEEE Transactions on Image Proceeding,2006,15(6): 1610-1620. |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|