Study On Light Detection Of Insoluble Particles In Glucose Injection

The insoluble particulate is an important detection index of pharmacopoeia specifications. Medical security problem caused by the insoluble particulate in injection has become a matter of urgent concern. Due to the limitation of traditional artificial visual detection and light-blockage detection method, the test efficiency of insoluble particles in injection is low, which has negative effects on the quality control of injection. Therefore, combining light scattering theory with visual image processing technology for the detection of insoluble particulate has become a trend of quality detection technology.Based on field investigation and references research, regarding glucose injection as the research object. This paper completed the overall design of visual image detection system of insoluble particulate in glucose injection; analyzed their sources and movement track; scattered light sequence images of insoluble particles under a particular illumination were acquired by high resolution camera; realized preprocessing of image under-research and calibration of the effective detection area; built image background model and studied inhibition analysis of the background noise and random noise; conducted a tracking research on insoluble particulate in Glucose Injection by combining background difference and improved Mean Shift algorithm; a rotating E-stop experiment platform was instituted. Then segmentation, extraction and identification of the detection object in image sequence were achieved according to simulation experiments, eventually the corresponding results were obtained.Based on the test mechanism study and experimental test analysis, this paper completed concentration test of the insoluble particulate in glucose injection; excluded bubble interference during insoluble particulate detection; analyzed the characteristic signal of glass fragment, wood, fiber, powder residues and other insoluble particles in glucose injection; verified the feasibility and rationality of light scattering visualization image sequence detection system; confirmed the good application prospects of light scattering visualization detection technology; accumulated useful experience for the further study of medicine intelligent detection.